Cognitive Liberties
Public claim and source map

Evidence you can inspect.

Every section in the 48-report worldwide research library maps to a stable evidence unit. This page exposes what is supported, what remains editorial analysis, what is a proposal or scenario, which sources were reviewed, and where verification remains limited.

194 evidence units 1141 inventoried assertions 84 source records As of 2026-09-02

Direct answer

An evidence label describes the cited scope—not the truth of every sentence in a section. Primary records can verify a law or holding; original research can support a measured finding; first-party guidance can document a published rule. Comparisons, risk judgments, policy recommendations, and future scenarios remain visibly classified as analysis even when their premises are sourced.

What this ledger does

Maps the public textOne stable evidence unit covers each public research section, with every sentence, list item, table row, or callout inventoried inside it.
Separates kinds of supportPrimary law, official oversight, standards, platform guidance, original research, methodology, and secondary analysis are not treated as interchangeable.
Preserves uncertaintyBlocked sources, incomplete rechecks, jurisdiction limits, model-specific findings, and report-derived arguments remain explicit.
Schedules reviewEach source and evidence unit carries an as-of or next-review date so time-sensitive statements do not silently become timeless claims.
Important boundary: The public ledger is a claim-to-source map and audit aid. It is not a guarantee that every source is correct, that a provider behaves exactly as its documentation states, or that a legal rule applies outside its facts and jurisdiction.

Verification states

The site uses nine explicit states. Counts describe this release, not a score or ranking.

Verified primary record2

A primary legal, official, or oversight source directly supports the section’s central factual foundation.

Verified original research2

Original research directly supports the technical or empirical foundation, subject to the study’s scope and limits.

Verified official guidance2

An official standard or first-party guidance directly supports the described architecture or published rule, not all real-world behavior.

Corroborated secondary evidence2

Multiple relevant secondary or civil-society sources support the description, but the section is not independently established by a single primary record.

Evidence + editorial analysis77

Sources support important factual premises; the synthesis, comparison, risk judgment, or conclusion remains editorial analysis.

Report-derived argument47

The supplied report is the primary basis for the analytical framing; external sources provide context but do not prove the whole proposition.

Site policy proposal55

A CognitiveLiberties.com recommendation or test, not current law, provider policy, or settled consensus.

Scenario analysis5

A bounded future scenario used to test consequences, not a forecast or claim that the scenario will occur.

Verification limited2

The source identity is known, but retrieval or direct review was incomplete during this release; reliance should remain qualified.

Section-level claim map

Evidence units

Each card maps one visible article section to the sources that support its overall evidentiary foundation. The assertion inventory preserves the exact public propositions for review; source IDs should not be read as sentence-by-sentence proof unless the support summary says so.

4 evidence units
Evidence + editorial analysis

Liberty begins before expression

CL-01-01
As of
Next review
Claim kind
legal context

Section summary: Most familiar civil liberties protect an outward act: speaking, publishing, worshiping, assembling, voting, or refusing a search.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-OHCHR-ICCPR International Covenant on Civil and Political Rights — Article 18 · Primary law or court record
  • SRC-OHCHR-GC22 Human Rights Committee General Comment No. 22 · Primary law or court record
  • SRC-OHCHR-A76-380 UN Special Rapporteur report A/76/380 — Freedom of thought · Official oversight record
Inspect 8 inventoried assertions
  1. CL-01-01.01 Most familiar civil liberties protect an outward act: speaking, publishing, worshiping, assembling, voting, or refusing a search. descriptive assertion
  2. CL-01-01.02 Those acts begin inside a private process of attention, memory, imagination, doubt, comparison, and belief formation. descriptive assertion
  3. CL-01-01.03 International human-rights language often describes the protected inner sphere as the forum internum —the internal domain of thought and conscience. descriptive assertion
  4. CL-01-01.04 That inner domain was historically difficult to invade directly. descriptive assertion
  5. CL-01-01.05 The law could focus on speech and conduct because the mind appeared to be a naturally protected space. inference or risk
  6. CL-01-01.06 Digital systems change the practical assumption. descriptive assertion
  7. CL-01-01.07 A person’s searches, reading sequence, location history, purchases, messages, pauses, corrections, and AI conversations can be combined into a revealing model of interests, fears, affiliations, health concerns, or likely preferences. inference or risk
  8. CL-01-01.08 Even when an inference is wrong, it can still shape what the person is shown or how the person is treated. inference or risk
Evidence + editorial analysis

What has changed in the AI era

CL-01-02
As of
Next review
Claim kind
technical capability

Section summary: The change is not one device or one model.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 14 inventoried assertions
  1. CL-01-02.01 The change is not one device or one model. descriptive assertion
  2. CL-01-02.02 It is the convergence of identity systems, behavioral data, prediction, personalization, generative AI, and increasingly sensitive sensors. descriptive assertion
  3. CL-01-02.03 The same ecosystem may know who a person is, what they searched, what they read, what they asked an assistant, where they went, and what opportunities they later received. inference or risk
  4. CL-01-02.04 AI increases the significance of this data because it can derive higher-level patterns from records that once looked mundane. inference or risk
  5. CL-01-02.05 A query log may become a health or political inference. inference or risk
  6. CL-01-02.06 A stream of writing may become a personality estimate. inference or risk
  7. CL-01-02.07 A recommendation system may use those estimates to change the informational environment in real time. inference or risk
  8. CL-01-02.08 The risk is therefore both observational and interventionist: systems can model the person and then act on the model. inference or risk
  9. CL-01-02.09 Immediate and more speculative cognitive-liberty risks descriptive assertion
  10. CL-01-02.10 Immediate or already practical | More limited or speculative structured assertion
  11. CL-01-02.11 Identity-linked search and AI histories | Remote decoding of a complete unspoken inner monologue structured assertion
  12. CL-01-02.12 Behavioral and psychographic inference | Reliable universal mind reading without cooperation structured assertion
  13. CL-01-02.13 Engagement optimization and personalized persuasion | Instant rewriting of a person’s beliefs through a single device structured assertion
  14. CL-01-02.14 Workplace, education, insurance, or advertising profiles | Perfect prediction of future conduct from neural data structured assertion
Evidence + editorial analysis

Why existing rights may be insufficient on their own

CL-01-03
As of
Next review
Claim kind
legal analysis

Section summary: Existing rights supply important foundations.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 14 inventoried assertions
  1. CL-01-03.01 Existing rights supply important foundations. descriptive assertion
  2. CL-01-03.02 Freedom of thought protects the inner domain. descriptive assertion
  3. CL-01-03.03 Privacy law limits some collection and use. descriptive assertion
  4. CL-01-03.04 The First Amendment protects expression and, in important contexts, receipt of information. descriptive assertion
  5. CL-01-03.05 The Fourth and Fifth Amendments constrain state searches and compelled testimony. descriptive assertion
  6. CL-01-03.06 Bodily integrity limits non-consensual intervention. descriptive assertion
  7. CL-01-03.07 The gap appears when a practice falls between those categories. descriptive assertion
  8. CL-01-03.08 A platform may infer a sensitive trait without collecting a traditionally regulated datum. inference or risk
  9. CL-01-03.09 A search system may silently narrow discovery without formally banning speech. inference or risk
  10. CL-01-03.10 An employer may rely on a behavioral score rather than a medical record. inference or risk
  11. CL-01-03.11 A conversational system may influence a user without an obvious moment of coercion. inference or risk
  12. CL-01-03.12 “Cognitive liberty” is useful because it asks whether the combined effect preserves mental self-determination, not merely whether each individual data transaction passed a narrow compliance test. descriptive assertion
  13. CL-01-03.13 Critics reasonably warn against creating a new right whenever technology changes. descriptive assertion
  14. CL-01-03.14 The alternative is to treat cognitive liberty as an organizing doctrine: it can modernize and connect existing protections while leaving room for courts and legislatures to decide whether a distinct legal right is necessary. inference or risk
Site policy proposal

A minimum bill of cognitive liberties

CL-01-04
As of
Next review
Claim kind
policy proposal

Section summary: Mental privacy: no non-consensual extraction, retention, sale, or high-impact use of neural data or sensitive cognitive inferences.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-01-04.01 Mental privacy: no non-consensual extraction, retention, sale, or high-impact use of neural data or sensitive cognitive inferences. proposal
  2. CL-01-04.02 Freedom of inquiry: lawful reading, searching, questioning, and private exploration should not create a presumption of dangerousness. proposal
  3. CL-01-04.03 Protection from covert manipulation: systems should not secretly exploit psychological vulnerabilities to bypass reflective choice. proposal
  4. CL-01-04.04 Cognitive self-determination: people should be able to choose or refuse neurotechnology, AI memory, personalization, and other cognitive tools without disproportionate penalty. proposal
  5. CL-01-04.05 Epistemic due process: consequential automated restrictions on information or opportunities should provide notice, reasons, correction, and meaningful review. proposal
  6. CL-01-04.06 Source and identity integrity: original statements, memories, inferred profiles, and machine-generated derivatives should remain distinguishable and correctable. proposal
  7. CL-01-04.07 These principles do not eliminate safety, fraud prevention, or regulation of harmful conduct. proposal
  8. CL-01-04.08 They establish a burden: interventions should be specific, proportionate, reviewable, and designed so that the ordinary intellectual life of the public does not become the object of generalized surveillance. proposal
4 evidence units
Report-derived argument

The anatomy of a safety poison pill

CL-02-01
As of
Next review
Claim kind
institutional risk

Section summary: The pattern begins with a harm that is vivid, severe, and politically difficult to oppose: terrorism, child exploitation, catastrophic cyberattack, fraud, or an imminent threat to life.

Support scope: The supplied source report is the main basis for this analytical framing. Public sources provide context but do not independently prove the full proposition.

Limit: The framing remains an argument from the supplied report and should be tested against additional primary evidence.

Conflict or tension: The external sources support parts of the framing but do not independently establish the complete report-derived argument.

Mapped sources
  • SRC-PCLOB-SECTION-215-2014 Privacy and Civil Liberties Oversight Board — Section 215 report · Official oversight record
  • SRC-DOJ-OIG-FBI-NSL U.S. Department of Justice Inspector General — FBI and National Security Letter reports · Official oversight record
Inspect 11 inventoried assertions
  1. CL-02-01.01 The pattern begins with a harm that is vivid, severe, and politically difficult to oppose: terrorism, child exploitation, catastrophic cyberattack, fraud, or an imminent threat to life. descriptive assertion
  2. CL-02-01.02 The proposed response then requires infrastructure capable of doing much more than addressing the original case. descriptive assertion
  3. CL-02-01.03 It may verify identity, retain every query, scan private content, assign behavioral risk, or create a centralized power to define what is “unsafe.” inference or risk
  4. CL-02-01.04 The public debate focuses on the horror of the predicate harm. descriptive assertion
  5. CL-02-01.05 The secondary capability receives less attention. descriptive assertion
  6. CL-02-01.06 Once deployed, the infrastructure gains sunk costs, vendors, institutional owners, and additional users. descriptive assertion
  7. CL-02-01.07 The definition of risk can expand while the technical mechanism remains in place. inference or risk
  8. CL-02-01.08 1. Salient harm A concrete or feared catastrophe dominates attention. descriptive assertion
  9. CL-02-01.09 2. Broad mechanism Monitoring or control is built for everyone, not only a target. descriptive assertion
  10. CL-02-01.10 3. Normalization The exceptional tool becomes routine infrastructure. inference or risk
  11. CL-02-01.11 4. Secondary use New agencies, offenses, or policies gain access. descriptive assertion
Evidence + editorial analysis

Where the thesis survives adversarial scrutiny

CL-02-02
As of
Next review
Claim kind
historical and empirical analysis

Section summary: The historical record strongly supports several narrower claims.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-PCLOB-SECTION-215-2014 Privacy and Civil Liberties Oversight Board — Section 215 report · Official oversight record
  • SRC-DOJ-OIG-FBI-NSL U.S. Department of Justice Inspector General — FBI and National Security Letter reports · Official oversight record
  • SRC-PENNEY-WIKIPEDIA-CHILL Jon Penney — Chilling Effects: Online Surveillance and Wikipedia Use · Original peer-reviewed research
  • SRC-MARTHEWS-TUCKER-SEARCH-CHILL Marthews and Tucker — Government Surveillance and Internet Search Behavior · Original preprint or working paper
Inspect 8 inventoried assertions
  1. CL-02-02.01 The historical record strongly supports several narrower claims. descriptive assertion
  2. CL-02-02.02 Exceptional threats can make broad authority easier to enact than to unwind. inference or risk
  3. CL-02-02.03 Secret or technically opaque systems are difficult to challenge. descriptive assertion
  4. CL-02-02.04 Bulk collection can persist before its specific value is proven. inference or risk
  5. CL-02-02.05 People can change what they read or search when they believe intellectual activity is monitored. inference or risk
  6. CL-02-02.06 Broad labels such as “dangerous,” “extremist,” or “high risk” create opportunities for mission creep when the decision-maker, evidence threshold, and appeal route are undefined. descriptive assertion
  7. CL-02-02.07 The thesis is particularly strong when a measure links identity to a person’s private inquiry and allows that history to influence unrelated decisions. descriptive assertion
  8. CL-02-02.08 A system that blocks a malicious transaction is different from one that records every controversial question and later treats the questioner as a risk. descriptive assertion
Evidence + editorial analysis

Where the thesis must be narrowed

CL-02-03
As of
Next review
Claim kind
comparative safety analysis

Section summary: Some safety interventions are specific enough that the poison-pill analogy does not fit.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 6 inventoried assertions
  1. CL-02-03.01 Some safety interventions are specific enough that the poison-pill analogy does not fit. descriptive assertion
  2. CL-02-03.02 Warning a user before a known malicious website, comparing a file against a database of known illegal imagery, requiring stronger authentication for a suspicious payment, or refusing narrowly operational assistance that would materially enable catastrophic harm can target conduct or capability without constructing a general dossier of thought. inference or risk
  3. CL-02-03.03 Age assurance also illustrates why architecture matters. inference or risk
  4. CL-02-03.04 A system that requires every reader to upload government identification creates a much greater cognitive-liberty risk than a token that proves only “over 18” and cannot be reused to track browsing. inference or risk
  5. CL-02-03.05 The relevant question is not whether a safeguard exists, but what information and power the safeguard creates. descriptive assertion
  6. CL-02-03.06 Adversarial verdict: “All safety is control” fails. “Safety claims require proof and must not become a shortcut to population-scale cognitive surveillance” survives strongly. descriptive assertion
Site policy proposal

The anti-poison-pill test

CL-02-04
As of
Next review
Claim kind
evaluation framework

Section summary: Questions that determine whether a safety proposal protects or controls

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 10 inventoried assertions
  1. CL-02-04.01 Questions that determine whether a safety proposal protects or controls proposal
  2. CL-02-04.02 Gate | Required answer | Failure signal proposal
  3. CL-02-04.03 Specific harm | The harm, probability, severity, and affected population are defined. | Vague appeals to “harm,” “misinformation,” or “risk.” proposal
  4. CL-02-04.04 Necessity | Evidence shows the measure materially reduces that harm. | Activity metrics are substituted for outcome evidence. proposal
  5. CL-02-04.05 Least restriction | Warnings, local controls, or targeted enforcement were considered first. | Universal logging or identity verification is the default. proposal
  6. CL-02-04.06 Purpose limits | Secondary use is technically and legally blocked. | Data may be repurposed for unrelated policing or profiling. proposal
  7. CL-02-04.07 Due process | People receive notice, reasons, correction, and appeal. | Secret classification or unreviewable automated denial. proposal
  8. CL-02-04.08 Reversibility | Hard sunsets, deletion, and dismantling are mandatory. | The authority or data persists indefinitely. proposal
  9. CL-02-04.09 A proposal that cannot answer these questions should remain unbuilt. proposal
  10. CL-02-04.10 The burden belongs to the actor seeking power over private inquiry, not to the public asked to surrender it. proposal
4 evidence units
Report-derived argument

Start with the harm, not the slogan

CL-03-01
As of
Next review
Claim kind
risk assessment method

Section summary: A useful safety proposal identifies the event it is designed to prevent, who is at risk, how likely the event is, how severe the harm would be, and what evidence connects the intervention to the outcome.

Support scope: The supplied source report is the main basis for this analytical framing. Public sources provide context but do not independently prove the full proposition.

Limit: The framing remains an argument from the supplied report and should be tested against additional primary evidence.

Conflict or tension: The external sources support parts of the framing but do not independently establish the complete report-derived argument.

Mapped sources
Inspect 10 inventoried assertions
  1. CL-03-01.01 A useful safety proposal identifies the event it is designed to prevent, who is at risk, how likely the event is, how severe the harm would be, and what evidence connects the intervention to the outcome. inference or risk
  2. CL-03-01.02 “Protect children,” “stop extremism,” or “prevent misinformation” names a goal, not an operational standard. descriptive assertion
  3. CL-03-01.03 The same analysis must identify uncertainty. descriptive assertion
  4. CL-03-01.04 A highly probable modest harm and a remote catastrophic harm create different policy questions. descriptive assertion
  5. CL-03-01.05 Institutions often blur them by emphasizing severity while omitting probability, or by citing a general problem without proving that the proposed system reduces it. descriptive assertion
  6. CL-03-01.06 What exact event counts as failure? structured assertion
  7. CL-03-01.07 What evidence shows it occurs at a meaningful rate? structured assertion
  8. CL-03-01.08 What population is affected? structured assertion
  9. CL-03-01.09 What measurable outcome should improve? structured assertion
  10. CL-03-01.10 What result would prove the intervention ineffective? structured assertion
Site policy proposal

Six gates for a liberty-preserving restriction

CL-03-02
As of
Next review
Claim kind
evaluation framework

Section summary: Specificity: the prohibited conduct or capability is objectively defined.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-03-02.01 Specificity: the prohibited conduct or capability is objectively defined. proposal
  2. CL-03-02.02 Necessity: evidence demonstrates a real contribution to reducing the harm. proposal
  3. CL-03-02.03 Proportionality: the intrusion is commensurate with the risk and does not sweep in large amounts of lawful inquiry. proposal
  4. CL-03-02.04 Alternatives: warnings, user controls, targeted investigations, rate limits, or transaction-level safeguards were considered first. proposal
  5. CL-03-02.05 Accountability: rules, error rates, enforcement statistics, and appeals are visible to independent review. proposal
  6. CL-03-02.06 Reversibility: data expires, authority sunsets, and infrastructure can actually be dismantled. proposal
  7. CL-03-02.07 A proposal that passes only the first gate is not ready. proposal
  8. CL-03-02.08 The existence of harm does not establish that a particular intervention is necessary or that a broad architecture is the least restrictive means. proposal
Evidence + editorial analysis

Distinguish protection from adjacent institutional motives

CL-03-03
As of
Next review
Claim kind
institutional incentive analysis

Section summary: Different motives that may be described as safety

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 10 inventoried assertions
  1. CL-03-03.01 Different motives that may be described as safety inference or risk
  2. CL-03-03.02 Category | Primary objective | Cognitive-liberty warning structured assertion
  3. CL-03-03.03 Genuine safety engineering | Reduce a defined, measurable harm. | Still requires evidence and narrow design. structured assertion
  4. CL-03-03.04 Paternalism | Override a person “for their own good.” | Adult autonomy and lawful inquiry are treated as incompetence. structured assertion
  5. CL-03-03.05 Liability management | Minimize legal exposure. | Risk-averse overblocking may replace accurate decisions. inference or risk
  6. CL-03-03.06 Reputational protection | Avoid controversy or advertiser pressure. | Unpopular ideas are reframed as user danger. structured assertion
  7. CL-03-03.07 Political censorship | Suppress opposition or inconvenient facts. | Vague safety labels track viewpoint rather than conduct. structured assertion
  8. CL-03-03.08 Institutional self-protection | Avoid scrutiny of the institution itself. | Whistleblowing and accountability research become “security threats.” structured assertion
  9. CL-03-03.09 These categories can overlap. inference or risk
  10. CL-03-03.10 The point is not to read minds but to examine incentives, decision rules, and measurable effects. descriptive assertion
Site policy proposal

A decision rule for AI and information systems

CL-03-04
As of
Next review
Claim kind
ai safety boundary

Section summary: For AI systems, the safest liberty-preserving sequence is usually: provide information; add context or warnings where risk is plausible; narrow the response when operational detail materially increases capability; refuse only the specific enabling portion; and escalate outside the interaction only when there is a lawful, high-confidence basis tied to an imminent or clearly defined threat.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 3 inventoried assertions
  1. CL-03-04.01 For AI systems, the safest liberty-preserving sequence is usually: provide information; add context or warnings where risk is plausible; narrow the response when operational detail materially increases capability; refuse only the specific enabling portion; and escalate outside the interaction only when there is a lawful, high-confidence basis tied to an imminent or clearly defined threat. proposal
  2. CL-03-04.02 This sequence protects a distinction that broad classifiers often erase: reading about a dangerous subject is not the same as advocating it, and advocacy is not the same as receiving interactive assistance that materially facilitates a harmful act. proposal
  3. CL-03-04.03 Rule: Safety should attach to demonstrable harmful conduct, a narrowly defined transaction, or a measurable capability uplift—not to the mere presence of a controversial topic in a person’s private inquiry. proposal
4 evidence units
Verified primary record

The post-9/11 pattern

CL-04-01
As of
Next review
Claim kind
historical oversight

Section summary: The September 11 attacks created an urgent and legitimate demand for better intelligence and faster coordination.

Support scope: The cited primary legal or official record directly supports the central legal or historical foundation in this section.

Limit: Application beyond the source’s facts, jurisdiction, or stated scope is an analogy and requires separate legal analysis.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-PCLOB-SECTION-215-2014 Privacy and Civil Liberties Oversight Board — Section 215 report · Official oversight record
  • SRC-DOJ-OIG-FBI-NSL U.S. Department of Justice Inspector General — FBI and National Security Letter reports · Official oversight record
Inspect 8 inventoried assertions
  1. CL-04-01.01 The September 11 attacks created an urgent and legitimate demand for better intelligence and faster coordination. descriptive assertion
  2. CL-04-01.02 The legal response also expanded surveillance authority quickly, used broad statutory language, and relied heavily on secret proceedings. descriptive assertion
  3. CL-04-01.03 Section 215 became the clearest example: language authorizing access to records “relevant” to an investigation was interpreted to support bulk telephone-metadata collection. descriptive assertion
  4. CL-04-01.04 National Security Letters added another layer. descriptive assertion
  5. CL-04-01.05 They enabled administrative demands for records and often carried non-disclosure requirements. descriptive assertion
  6. CL-04-01.06 Inspector-general reviews later documented compliance problems, errors, and misuse. descriptive assertion
  7. CL-04-01.07 Those failures do not require a theory of coordinated bad faith. descriptive assertion
  8. CL-04-01.08 They show that broad authority, secrecy, complex rules, and weak controls can produce rights violations through ordinary institutional behavior. inference or risk
Evidence + editorial analysis

The supported parallels to AI governance

CL-04-02
As of
Next review
Claim kind
historical analogy

Section summary: Supported institutional similarities

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-04-02.01 Supported institutional similarities descriptive assertion
  2. CL-04-02.02 Post-9/11 pattern | Possible AI-era equivalent | Risk structured assertion
  3. CL-04-02.03 Emergency enactment | Rapid rules justified by catastrophic model misuse | Insufficient analysis of long-term civil-liberties costs structured assertion
  4. CL-04-02.04 Broad predicates | “Unsafe,” “dual use,” “systemic risk,” or “extremist” | Scope can expand without new legislation inference or risk
  5. CL-04-02.05 Secret interpretation | Opaque classifiers, undisclosed thresholds, private government-platform channels | People cannot understand or challenge the rule structured assertion
  6. CL-04-02.06 Third-party access | Government access to platform prompt and search histories | Private cognitive records become investigative databases structured assertion
  7. CL-04-02.07 Mission creep | Biosecurity or cyber controls reused for routine offenses or political monitoring | Exceptional safety becomes general social control inference or risk
Report-derived argument

Where the analogy breaks

CL-04-03
As of
Next review
Claim kind
technical and legal distinction

Section summary: A telephone network transmits communications.

Support scope: The supplied source report is the main basis for this analytical framing. Public sources provide context but do not independently prove the full proposition.

Limit: The framing remains an argument from the supplied report and should be tested against additional primary evidence.

Conflict or tension: The external sources support parts of the framing but do not independently establish the complete report-derived argument.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-04-03.01 A telephone network transmits communications. descriptive assertion
  2. CL-04-03.02 A frontier AI model can also generate plans, code, designs, persuasive content, or troubleshooting assistance. inference or risk
  3. CL-04-03.03 That difference matters. descriptive assertion
  4. CL-04-03.04 If a model provides a user with a meaningful operational uplift toward a catastrophic biological or cyber harm, a request-level refusal may regulate capability rather than thought. inference or risk
  5. CL-04-03.05 The analogy therefore does not prove that all model safeguards are illegitimate. descriptive assertion
  6. CL-04-03.06 It shows why a narrow capability restriction should not automatically become identity-linked query surveillance. descriptive assertion
  7. CL-04-03.07 A model can decline a specific enabling step without storing a permanent psychological dossier on everyone who asked about the subject. inference or risk
Site policy proposal

Safeguards that should be built before the crisis

CL-04-04
As of
Next review
Claim kind
institutional safeguards

Section summary: Warrant protection for AI histories: prompts and assistant conversations should not lose privacy merely because a provider processed them.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-04-04.01 Warrant protection for AI histories: prompts and assistant conversations should not lose privacy merely because a provider processed them. proposal
  2. CL-04-04.02 Identity separation: verify high-risk transactions or compute purchases without linking ordinary end-user inquiry to legal identity. proposal
  3. CL-04-04.03 Minimization and deletion: retain only what is necessary for the immediate function and erase it on a fixed schedule. proposal
  4. CL-04-04.04 Adversarial oversight: secret or specialized proceedings need a cleared advocate whose job is to contest the government’s interpretation. proposal
  5. CL-04-04.05 Public statistics: publish request counts, error rates, purposes, and outcomes without exposing legitimate investigations. proposal
  6. CL-04-04.06 Hard sunsets: exceptional authority should terminate unless lawmakers publicly re-establish necessity. proposal
  7. CL-04-04.07 Decentralization: protect local and open systems so no single provider becomes a universal cognitive chokepoint. proposal
  8. CL-04-04.08 The central lesson is temporal: civil-liberties architecture must be designed before a crisis makes deliberation politically impossible. proposal
4 evidence units
Report-derived argument

The actor map

CL-05-01
As of
Next review
Claim kind
actor taxonomy

Section summary: Actors, powers, and primary cognitive-liberty risks

Support scope: The supplied source report is the main basis for this analytical framing. Public sources provide context but do not independently prove the full proposition.

Limit: The framing remains an argument from the supplied report and should be tested against additional primary evidence.

Conflict or tension: The external sources support parts of the framing but do not independently establish the complete report-derived argument.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-05-01.01 Actors, powers, and primary cognitive-liberty risks descriptive assertion
  2. CL-05-01.02 Actor | Power | Primary risk structured assertion
  3. CL-05-01.03 Legislatures and regulators | Define legal duties, access rules, identity requirements, and penalties | Broad mandates that normalize monitoring or censorship structured assertion
  4. CL-05-01.04 Law enforcement and intelligence agencies | Compel records, investigate, retain intelligence, and classify threats | Mission creep from exceptional targets to ordinary inquiry structured assertion
  5. CL-05-01.05 Search and social platforms | Rank, recommend, deindex, personalize, and moderate | Invisible constraints on what people encounter structured assertion
  6. CL-05-01.06 AI providers | Log prompts, infer intent, retain memory, refuse, and generate answers | Private judgment over questions and cognitive context structured assertion
  7. CL-05-01.07 Cloud, app stores, telecoms, and payment systems | Control hosting, distribution, connectivity, and financial viability | Infrastructure denial without meaningful appeal structured assertion
  8. CL-05-01.08 Data brokers and ad networks | Link identities, behaviors, locations, and inferred traits | Hidden cognitive profiles sold or repurposed structured assertion
  9. CL-05-01.09 Employers, schools, and universities | Condition opportunity on monitoring and behavioral scores | Coercive surveillance where consent is not freely given structured assertion
Evidence + editorial analysis

Mechanisms matter more than labels

CL-05-02
As of
Next review
Claim kind
evidence method

Section summary: The same institution can protect cognitive liberty in one context and threaten it in another.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-05-02.01 The same institution can protect cognitive liberty in one context and threaten it in another. inference or risk
  2. CL-05-02.02 A search company can warn users about malware while also retaining sensitive query histories. inference or risk
  3. CL-05-02.03 A school can protect students from harassment while deploying software that monitors every search. inference or risk
  4. CL-05-02.04 A government can enforce privacy law while seeking broad access to private records. inference or risk
  5. CL-05-02.05 Analysis should therefore focus on mechanisms: filtering, ranking, identity linkage, query retention, compelled disclosure, behavioral scoring, account sanctions, infrastructure denial, financial pressure, biometric extraction, and covert recommendation manipulation. descriptive assertion
  6. CL-05-02.06 Name an actor only at the evidence level the record supports. descriptive assertion
  7. CL-05-02.07 A demonstrated capability is not proof of deployment; an incentive is not proof of misconduct; a contested allegation is not an adjudicated fact. descriptive assertion
Report-derived argument

Which mechanisms present the greatest risk?

CL-05-03
As of
Next review
Claim kind
risk ranking

Section summary: Risk increases when five conditions converge: the mechanism reaches many people; the affected person cannot see it; the effect is difficult to reverse; there is no meaningful appeal; and the decision changes access to knowledge or opportunity.

Support scope: The supplied source report is the main basis for this analytical framing. Public sources provide context but do not independently prove the full proposition.

Limit: The framing remains an argument from the supplied report and should be tested against additional primary evidence.

Conflict or tension: The external sources support parts of the framing but do not independently establish the complete report-derived argument.

Mapped sources
Inspect 1 inventoried assertions
  1. CL-05-03.01 Risk increases when five conditions converge: the mechanism reaches many people; the affected person cannot see it; the effect is difficult to reverse; there is no meaningful appeal; and the decision changes access to knowledge or opportunity. descriptive assertion
Site policy proposal

A layered defense

CL-05-04
As of
Next review
Claim kind
layered defense

Section summary: No single law can govern the entire stack.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-05-04.01 No single law can govern the entire stack. proposal
  2. CL-05-04.02 Government action requires constitutional limits, warrants, transparency, and adversarial review. proposal
  3. CL-05-04.03 Dominant platforms require contestability, reason notices, portability, and competition. proposal
  4. CL-05-04.04 Data brokers require strict purpose limits and rights over inferred data. proposal
  5. CL-05-04.05 Employers and schools require heightened consent standards and categorical limits on emotion or attention surveillance. proposal
  6. CL-05-04.06 Infrastructure providers require clear, viewpoint-neutral terms and fair process before termination. proposal
  7. CL-05-04.07 Technical architecture is also a defense. proposal
  8. CL-05-04.08 Local processing, short retention, anonymous credentials, interoperable protocols, and decentralized publishing reduce the amount of cognitive power any one actor can accumulate. proposal
4 evidence units
Evidence + editorial analysis

Why queries are unusually sensitive

CL-06-01
As of
Next review
Claim kind
privacy analysis

Section summary: Published speech is normally composed for an audience.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 6 inventoried assertions
  1. CL-06-01.01 Published speech is normally composed for an audience. descriptive assertion
  2. CL-06-01.02 A search query may be a fragment, a fear, an experiment, a symptom, or a question the person is embarrassed to ask another human. inference or risk
  3. CL-06-01.03 It can contain a name, address, diagnosis, political idea, or imagined scenario without revealing why the person typed it. scenario
  4. CL-06-01.04 Sequences of queries are more revealing than a single entry. descriptive assertion
  5. CL-06-01.05 Timing, reformulation, frequency, and the path from one topic to another can support inferences that the user never explicitly disclosed. inference or risk
  6. CL-06-01.06 AI makes this aggregation more consequential because it can classify patterns across large histories and connect them with other behavioral data. inference or risk
Evidence + editorial analysis

How a query becomes a dossier

CL-06-02
As of
Next review
Claim kind
data flow analysis

Section summary: The privacy risk is produced by linkage.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-06-02.01 The privacy risk is produced by linkage. inference or risk
  2. CL-06-02.02 A provider may associate a query with an account, cookie, device, IP address, location, session, or advertising identifier. inference or risk
  3. CL-06-02.03 Even after direct identifiers are removed, the semantic content can reveal identity. inference or risk
  4. CL-06-02.04 The history may then be used for personalization, advertising, fraud detection, model improvement, legal compliance, or other secondary purposes. inference or risk
  5. CL-06-02.05 Each stage increases the danger of context collapse: a question asked for research may be treated as a trait, and a trait may become a decision. inference or risk
  6. CL-06-02.06 Query A private or tentative question. descriptive assertion
  7. CL-06-02.07 Association Account, IP, device, location, or cookie. descriptive assertion
  8. CL-06-02.08 Inference Health, ideology, intent, risk, or susceptibility. descriptive assertion
  9. CL-06-02.09 Consequence Ranking, ads, investigation, pricing, or eligibility. descriptive assertion
Verified primary record

The legal mismatch

CL-06-03
As of
Next review
Claim kind
legal analysis

Section summary: Physical libraries have a long tradition of confidentiality because the freedom to read depends on not being judged for the books one explores.

Support scope: The cited primary legal or official record directly supports the central legal or historical foundation in this section.

Limit: Application beyond the source’s facts, jurisdiction, or stated scope is an analogy and requires separate legal analysis.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-06-03.01 Physical libraries have a long tradition of confidentiality because the freedom to read depends on not being judged for the books one explores. descriptive assertion
  2. CL-06-03.02 Digital search often receives weaker practical protection because the query passes through a company. descriptive assertion
  3. CL-06-03.03 The traditional third-party doctrine treats some data voluntarily given to a service provider as outside a reasonable expectation of privacy. descriptive assertion
  4. CL-06-03.04 Modern Supreme Court cases have questioned the fit of that doctrine for exhaustive digital records. descriptive assertion
  5. CL-06-03.05 Carpenter recognized that sensitive, comprehensive location history can remain constitutionally protected even though a company holds it. inference or risk
  6. CL-06-03.06 Search history presents a parallel argument: it is pervasive, revealing, and difficult to avoid while participating in contemporary life. descriptive assertion
  7. CL-06-03.07 Reverse keyword warrants sharpen the problem. descriptive assertion
  8. CL-06-03.08 Instead of obtaining records about a known person, the government asks a provider to identify everyone who searched for specified terms. descriptive assertion
  9. CL-06-03.09 That procedure can sweep lawful curiosity into an investigative pool before individualized suspicion exists. inference or risk
Site policy proposal

A proposed doctrine of search privacy

CL-06-04
As of
Next review
Claim kind
search privacy doctrine

Section summary: Search minimization: do not retain identity-linked queries beyond the immediate technical need without explicit opt-in.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-06-04.01 Search minimization: do not retain identity-linked queries beyond the immediate technical need without explicit opt-in. proposal
  2. CL-06-04.02 Warrant protection: government access to identifiable search histories should ordinarily require probable cause, particularity, and heightened care for expressive material. proposal
  3. CL-06-04.03 No character inference: lawful queries alone should not determine employability, insurability, credit, trustworthiness, or access to services. proposal
  4. CL-06-04.04 Secondary-use limits: search histories should not silently become advertising, model-training, or behavioral-risk data. proposal
  5. CL-06-04.05 Inference rights: people should be able to inspect, correct, and delete high-impact sensitive inferences derived from query histories. proposal
  6. CL-06-04.06 Reverse-search limits: keyword demands should face strict necessity, narrow time and term scope, independent review, minimization, and notice when lawful. proposal
  7. CL-06-04.07 Presumption: A lawful query is evidence of inquiry, not evidence of character or intent. proposal
4 evidence units
Verified original research

From words to an inferred need

CL-07-01
As of
Next review
Claim kind
information retrieval

Section summary: Older search systems relied heavily on lexical overlap and link analysis.

Support scope: The cited original research supports the described empirical or technical foundation; the broader civil-liberties implications remain editorial analysis.

Limit: Study results may be dataset-, model-, version-, or laboratory-specific and should not be generalized beyond the reported conditions.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-ARXIV-QUERY-INTENT-2020 Deep Search Query Intent Understanding · Original preprint or working paper
  • SRC-ARXIV-RAG-2020 Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks · Original peer-reviewed research
Inspect 7 inventoried assertions
  1. CL-07-01.01 Older search systems relied heavily on lexical overlap and link analysis. descriptive assertion
  2. CL-07-01.02 Contemporary systems combine exact matching with transformer models, dense embeddings, entity recognition, and learned ranking. descriptive assertion
  3. CL-07-01.03 The query becomes a vector that can be compared with semantically related passages even when the same words do not appear. inference or risk
  4. CL-07-01.04 The system may classify the request as informational, navigational, transactional, local, medical, time-sensitive, or another intent class. inference or risk
  5. CL-07-01.05 It may also reformulate the query, correct spelling, expand concepts, or break a complex request into sub-queries. inference or risk
  6. CL-07-01.06 These interventions can be helpful, but they mean that the machine is no longer a neutral conduit for the literal text. inference or risk
  7. CL-07-01.07 It is interpreting what the user “really” wants. descriptive assertion
Evidence + editorial analysis

Where interpretation becomes profiling

CL-07-02
As of
Next review
Claim kind
profiling boundary

Section summary: Layers of query interpretation

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-07-02.01 Layers of query interpretation descriptive assertion
  2. CL-07-02.02 Layer | Useful purpose | Cognitive-liberty risk structured assertion
  3. CL-07-02.03 Language and location | Return locally and linguistically relevant results | Persistent geolocation or identity association structured assertion
  4. CL-07-02.04 Session context | Resolve ambiguous follow-up questions | Retaining sensitive topic sequences beyond the session structured assertion
  5. CL-07-02.05 Personalization | Reduce irrelevant results | Filter bubbles and hidden psychographic profiles structured assertion
  6. CL-07-02.06 Safety classification | Limit direct facilitation of serious harm | Conflating a topic with malicious intent structured assertion
  7. CL-07-02.07 Age estimation | Apply age-appropriate experiences | Behavioral surveillance or universal identity verification structured assertion
  8. CL-07-02.08 The decisive line is persistence and consequence. descriptive assertion
  9. CL-07-02.09 A temporary, request-local interpretation used only to answer the question is less intrusive than a durable “risk” label tied to identity and reused across services. descriptive assertion
Verified original research

Curiosity is difficult to classify

CL-07-03
As of
Next review
Claim kind
over refusal

Section summary: Intent is underdetermined by language.

Support scope: The cited original research supports the described empirical or technical foundation; the broader civil-liberties implications remain editorial analysis.

Limit: Study results may be dataset-, model-, version-, or laboratory-specific and should not be generalized beyond the reported conditions.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-07-03.01 Intent is underdetermined by language. descriptive assertion
  2. CL-07-03.02 “How does ransomware spread?” may be defensive security research, journalism, education, or malicious planning. inference or risk
  3. CL-07-03.03 A system can often identify the topic and the operational specificity of a request, but it cannot reliably know the user’s complete motive from one text string. inference or risk
  4. CL-07-03.04 Safety tuning therefore produces false positives. descriptive assertion
  5. CL-07-03.05 Models may refuse historical analysis, fiction, policy research, or defensive explanations because the vocabulary resembles a prohibited request. inference or risk
  6. CL-07-03.06 The error is not harmless. descriptive assertion
  7. CL-07-03.07 It teaches users that some questions are suspect, increases the advantage of technically sophisticated users who know how to reframe prompts, and can deny access to people seeking help or conducting legitimate work. inference or risk
Site policy proposal

Epistemic due process for search

CL-07-04
As of
Next review
Claim kind
epistemic due process

Section summary: Notice: disclose when safety, personalization, age, or other classification materially altered the result.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-07-04.01 Notice: disclose when safety, personalization, age, or other classification materially altered the result. proposal
  2. CL-07-04.02 Reason: identify the rule or category in plain language without exposing exploitable security details. proposal
  3. CL-07-04.03 Context correction: let the user clarify research, journalistic, legal, fictional, or defensive context. proposal
  4. CL-07-04.04 Appeal: provide human review for repeated or consequential denials. proposal
  5. CL-07-04.05 Profile access: allow inspection and deletion of persistent sensitive inferences. proposal
  6. CL-07-04.06 Ephemerality: discard request-level intent and safety labels when the session ends unless retention is truly necessary. proposal
  7. CL-07-04.07 Identity separation: use local processing, anonymous routing, or isolated classifiers so safety evaluation does not become a personal dossier. proposal
  8. CL-07-04.08 A trustworthy system can interpret language without judging the whole person. proposal
  9. CL-07-04.09 It should answer the request, disclose its boundary, and forget the temporary classification. proposal
4 evidence units
Verified official guidance

What classical search evaluates

CL-08-01
As of
Next review
Claim kind
search quality

Section summary: A useful search engine must crawl pages, understand their content, identify duplicates, resist manipulation, and rank results for a particular query.

Support scope: The cited official standard or first-party guidance supports the published architecture or rule, but does not prove every production implementation.

Limit: Published guidance can change and does not guarantee ranking, citation, compliance, or identical behavior in every deployment.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 6 inventoried assertions
  1. CL-08-01.01 A useful search engine must crawl pages, understand their content, identify duplicates, resist manipulation, and rank results for a particular query. descriptive assertion
  2. CL-08-01.02 Publicly documented considerations include relevance, technical accessibility, security, originality, link relationships, freshness when the query requires it, and systems designed to detect spam or deceptive behavior. descriptive assertion
  3. CL-08-01.03 These systems are not optional. descriptive assertion
  4. CL-08-01.04 Without them, results would be dominated by cloaking, link schemes, scraped pages, malware, and scaled low-value content. descriptive assertion
  5. CL-08-01.05 Cognitive liberty does not require a search engine to rank every page equally. descriptive assertion
  6. CL-08-01.06 It requires that the gatekeeping power be understandable, contestable in high-impact cases, and not silently repurposed into viewpoint control. descriptive assertion
Evidence + editorial analysis

What generative discovery adds

CL-08-02
As of
Next review
Claim kind
generative retrieval

Section summary: AI-mediated search changes the user experience from choosing among documents to receiving a synthesized answer.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-ARXIV-RAG-2020 Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks · Original peer-reviewed research
  • SRC-GOOGLE-AI-OPTIMIZATION Google guide to optimizing for generative AI features · First-party platform guidance
Inspect 10 inventoried assertions
  1. CL-08-02.01 AI-mediated search changes the user experience from choosing among documents to receiving a synthesized answer. descriptive assertion
  2. CL-08-02.02 A retrieval pipeline may embed the query, retrieve semantically related passages, rerank them, select sources, generate prose, and attach citations. inference or risk
  3. CL-08-02.03 Every step can exclude a source before the user ever sees it. inference or risk
  4. CL-08-02.04 Generative answers increase convenience but reduce source diversity by design. descriptive assertion
  5. CL-08-02.05 The system must choose a limited evidence set and compress disagreement. descriptive assertion
  6. CL-08-02.06 If citation and uncertainty are weak, the user may receive a confident answer that hides contested evidence or minority analysis. inference or risk
  7. CL-08-02.07 Crawl and index Can the system access and parse the page? inference or risk
  8. CL-08-02.08 Retrieve Does a passage match the query semantically or lexically? descriptive assertion
  9. CL-08-02.09 Rerank Which sources appear most relevant and trustworthy? descriptive assertion
  10. CL-08-02.10 Synthesize Which claims survive into the generated answer? descriptive assertion
Evidence + editorial analysis

The heterodoxy and small-publisher problem

CL-08-03
As of
Next review
Claim kind
discoverability risk

Section summary: Authority signals often rely on history, third-party references, consistent identity, and recognized expertise.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-08-03.01 Authority signals often rely on history, third-party references, consistent identity, and recognized expertise. descriptive assertion
  2. CL-08-03.02 Those signals help filter scams, but they also reward incumbency. descriptive assertion
  3. CL-08-03.03 A new publisher cannot begin with years of citations. descriptive assertion
  4. CL-08-03.04 A whistleblower may need anonymity. inference or risk
  5. CL-08-03.05 A specialized researcher may challenge current consensus precisely because the prevailing institutions are incomplete or wrong. inference or risk
  6. CL-08-03.06 The solution is not to discard quality systems. descriptive assertion
  7. CL-08-03.07 It is to keep factual reliability distinct from institutional conformity, protect anonymous evidence when independently verifiable, and offer corrective process when a domain-level classifier suppresses an entire body of work. descriptive assertion
  8. CL-08-03.08 Discoverability principle: Ranking may reflect relevance and evidence quality; it should not become an unreviewable punishment for lawful viewpoint, anonymity, novelty, or lack of institutional status. inference or risk
Site policy proposal

Algorithmic Discoverability Due Process

CL-08-04
As of
Next review
Claim kind
publisher due process

Section summary: Proposed protections for publishers affected by automated gatekeeping

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 11 inventoried assertions
  1. CL-08-04.01 Proposed protections for publishers affected by automated gatekeeping proposal
  2. CL-08-04.02 Protection | What it would provide proposal
  3. CL-08-04.03 Demotion notice | Confirmation that a sustained domain- or page-level classifier materially affected visibility. proposal
  4. CL-08-04.04 Reason category | A standardized explanation such as malware, deceptive practice, scaled abuse, duplicate content, or source-trust failure. proposal
  5. CL-08-04.05 Evidence and examples | Representative affected URLs and actionable diagnostic detail. proposal
  6. CL-08-04.06 Human appeal | Review for cultural context, scholarship, public-interest reporting, and classifier error. proposal
  7. CL-08-04.07 Timely correction | Recovery after remediation without waiting indefinitely for a broad system update. proposal
  8. CL-08-04.08 Transparency reports | Aggregate data on classifier reach, error correction, and appeal outcomes. proposal
  9. CL-08-04.09 Viewpoint protection | Audits for systematic effects on lawful political, scientific, religious, or minority perspectives. proposal
  10. CL-08-04.10 For publishers, the durable strategy remains straightforward: make pages accessible, original, specific, well sourced, semantically clear, and honest about who produced them. proposal
  11. CL-08-04.11 Structured data can clarify entities and relationships, but no markup can substitute for evidence or guarantee AI citation. proposal
4 evidence units
Evidence + editorial analysis

Why difficult research is socially necessary

CL-09-01
As of
Next review
Claim kind
professional and social need

Section summary: Preventing harm often requires direct study of the thing that causes it.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-09-01.01 Preventing harm often requires direct study of the thing that causes it. descriptive assertion
  2. CL-09-01.02 Counter-extremism researchers read propaganda. descriptive assertion
  3. CL-09-01.03 Cybersecurity professionals examine malware and exploit chains. descriptive assertion
  4. CL-09-01.04 Public-health experts study suicide and drug use. descriptive assertion
  5. CL-09-01.05 Historians preserve records of genocide and authoritarianism. descriptive assertion
  6. CL-09-01.06 Attorneys investigate alleged crimes. descriptive assertion
  7. CL-09-01.07 Journalists reconstruct trafficking, corruption, and violence. descriptive assertion
  8. CL-09-01.08 Broad filters can erase these distinctions because the same words occur in defensive and malicious contexts. inference or risk
  9. CL-09-01.09 The result is an information asymmetry: sophisticated bad actors continue to seek knowledge while ordinary users, students, and defenders lose access or fear creating a suspicious record. descriptive assertion
Site policy proposal

A seven-level inquiry taxonomy

CL-09-02
As of
Next review
Claim kind
intent taxonomy

Section summary: From curiosity to imminent action

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 11 inventoried assertions
  1. CL-09-02.01 From curiosity to imminent action proposal
  2. CL-09-02.02 Level | Description | Default response proposal
  3. CL-09-02.03 1. Curiosity | Unstructured interest, fiction, or cultural exploration | Answer fully proposal
  4. CL-09-02.04 2. Education | Historical, scientific, legal, or conceptual learning | Answer fully with sources proposal
  5. CL-09-02.05 3. Analysis | Journalism, research, criticism, prevention, or defense | Answer fully; preserve professional context proposal
  6. CL-09-02.06 4. Advocacy | Argument for or against a controversial idea | Protect lawful expression; do not infer action proposal
  7. CL-09-02.07 5. Preparation | Dual-use materials or general planning without a target | Assess specificity and capability uplift carefully proposal
  8. CL-09-02.08 6. Operational facilitation | Interactive steps that materially overcome barriers to a harmful act | Narrow or refuse the enabling assistance proposal
  9. CL-09-02.09 7. Imminent harmful action | Specific target, timing, means, and immediate intent | Refuse and follow lawful, narrowly defined emergency process proposal
  10. CL-09-02.10 The taxonomy is not a mind-reading machine. proposal
  11. CL-09-02.11 It is a discipline for focusing on what the requested assistance does rather than what topic it mentions. proposal
Evidence + editorial analysis

Capability uplift is a better boundary

CL-09-03
As of
Next review
Claim kind
capability boundary

Section summary: A high-level explanation of how a pathogen spreads may provide no meaningful advantage beyond public sources.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-09-03.01 A high-level explanation of how a pathogen spreads may provide no meaningful advantage beyond public sources. inference or risk
  2. CL-09-03.02 Interactive troubleshooting that helps evade screening or optimize a dangerous process can provide a large advantage. inference or risk
  3. CL-09-03.03 The latter changes the user’s practical capacity, which supplies a stronger safety justification than the former. descriptive assertion
  4. CL-09-03.04 The same logic applies to cyber research. descriptive assertion
  5. CL-09-03.05 Explaining a vulnerability class supports defense. descriptive assertion
  6. CL-09-03.06 Writing or debugging a tailored payload against a named hospital system crosses into operational facilitation. descriptive assertion
  7. CL-09-03.07 A precise system can preserve the explanation while refusing the targeted component. inference or risk
Evidence + editorial analysis

Presumption, privacy, and evidence

CL-09-04
As of
Next review
Claim kind
legal and privacy analysis

Section summary: Query and browsing records are ambiguous evidence.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-09-04.01 Query and browsing records are ambiguous evidence. descriptive assertion
  2. CL-09-04.02 A disturbing search may be relevant in a criminal case when it is tightly connected to a charged act, specific method, target, and time. inference or risk
  3. CL-09-04.03 It should not become a general license to treat everyone who searched a term as a suspect. descriptive assertion
  4. CL-09-04.04 Reverse keyword warrants are especially risky because they begin with a term and ask a provider to identify unknown users. descriptive assertion
  5. CL-09-04.05 This can convert lawful inquiry into a dragnet. inference or risk
  6. CL-09-04.06 Strong safeguards require necessity, narrow scope, probable cause, minimization, independent review, and particular care for expressive material. descriptive assertion
  7. CL-09-04.07 For AI and search systems, uncertainty should favor continued analytical assistance. descriptive assertion
  8. CL-09-04.08 The system can offer history, prevention, ethics, legal context, and defensive information while withholding only the portion that materially enables a specific harm. inference or risk
4 evidence units
Evidence + editorial analysis

Change the architecture, not the promise

CL-10-01
As of
Next review
Claim kind
privacy architecture principle

Section summary: A privacy policy can promise that data will not be misused, but a retained database remains available to insiders, attackers, future owners, subpoenas, and changed policies.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 4 inventoried assertions
  1. CL-10-01.01 A privacy policy can promise that data will not be misused, but a retained database remains available to insiders, attackers, future owners, subpoenas, and changed policies. inference or risk
  2. CL-10-01.02 A stronger design does not collect the identifying data in the first place or separates it so no single party can reconstruct the whole interaction. inference or risk
  3. CL-10-01.03 This is the difference between institutional restraint and technical restraint. descriptive assertion
  4. CL-10-01.04 Both matter, but the latter reduces the number of future decisions that can go wrong. inference or risk
Verified official guidance

Five privacy-preserving building blocks

CL-10-02
As of
Next review
Claim kind
privacy enhancing technologies

Section summary: Privacy-enhancing technologies for safety functions

Support scope: The cited official standard or first-party guidance supports the published architecture or rule, but does not prove every production implementation.

Limit: Published guidance can change and does not guarantee ranking, citation, compliance, or identical behavior in every deployment.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-10-02.01 Privacy-enhancing technologies for safety functions descriptive assertion
  2. CL-10-02.02 Tool | What it protects | Example safety use structured assertion
  3. CL-10-02.03 On-device models | Raw prompts, messages, images, and behavioral data | Local content classification or PII redaction structured assertion
  4. CL-10-02.04 Oblivious HTTP | Links between IP identity and request content | Malicious-site checks without exposing browsing history structured assertion
  5. CL-10-02.05 Private Information Retrieval | The specific record requested from a database | Private lookup of a URL or file hash in a threat list structured assertion
  6. CL-10-02.06 Distributed aggregation + differential privacy | Individual telemetry and rare-event identification | Measure abuse trends without collecting per-user histories structured assertion
  7. CL-10-02.07 Anonymous credentials | Identity beyond the required attribute | Prove “adult” or “human” without revealing name or birthday structured assertion
Evidence + editorial analysis

Two concrete patterns

CL-10-03
As of
Next review
Claim kind
architecture patterns

Section summary: A browser hashes a URL, prepares a private lookup, and sends it through an oblivious relay.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 5 inventoried assertions
  1. CL-10-03.01 A browser hashes a URL, prepares a private lookup, and sends it through an oblivious relay. descriptive assertion
  2. CL-10-03.02 The threat provider can answer whether the resource is known to be dangerous without learning both the user’s identity and the exact browsing request. inference or risk
  3. CL-10-03.03 A local classifier identifies a safety category on the device. descriptive assertion
  4. CL-10-03.04 Instead of uploading the content, the device secret-shares a minimal event count across independent aggregators. descriptive assertion
  5. CL-10-03.05 The service learns that a category is increasing across the population but cannot reconstruct an individual person’s message. descriptive assertion
Evidence + editorial analysis

Trade-offs and limits

CL-10-04
As of
Next review
Claim kind
technical tradeoffs

Section summary: Privacy-preserving systems introduce costs.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 10 inventoried assertions
  1. CL-10-04.01 Privacy-preserving systems introduce costs. descriptive assertion
  2. CL-10-04.02 Relays add latency. descriptive assertion
  3. CL-10-04.03 Private database queries require additional computation. descriptive assertion
  4. CL-10-04.04 Local models consume battery and may perform worse on older devices. inference or risk
  5. CL-10-04.05 Multi-party systems depend on non-collusion or cryptographic assumptions. descriptive assertion
  6. CL-10-04.06 Differential privacy can reduce precision, especially for rare events. inference or risk
  7. CL-10-04.07 Those costs should be compared with the security cost of centralized data: breaches, insider access, government overreach, and function creep. descriptive assertion
  8. CL-10-04.08 The proper question is not whether privacy has zero cost. descriptive assertion
  9. CL-10-04.09 It is whether retaining a complete cognitive history is necessary for the safety outcome. descriptive assertion
  10. CL-10-04.10 Design rule: If the same safety result can be achieved without retaining a person’s inquiry history, the system should not retain that history. inference or risk
4 evidence units
Verification limited

Ten forces that drive expansion

CL-11-01
As of
Next review
Claim kind
function creep analysis

Section summary: Sunk costs Once a system exists, institutions seek more uses for it. Bureaucratic growth Budgets and relevance increase with mission scope. Secrecy Hidden rules reduce public and adversarial review. Broad definitions Elastic terms permit new targets without new authority. Capability Available tools create pressure to use them. Weak sunsets Renewal becomes routine and politically asymmetric. Political fear Officials are blamed for missed threats more than diffuse privacy loss. Normalization Repeated exposure lowers public resistance. Vendor incentives Suppliers lobby for continuing and expanded markets. Data utility Records collected for one purpose prove attractive for another.

Support scope: The cited source identity is preserved, but direct source retrieval was blocked or incomplete in this review; the section retains an explicit verification limit.

Limit: Do not rely on the section as fully reverified until the identified source is successfully reviewed.

Conflict or tension: Direct source retrieval remains incomplete.

Mapped sources
Inspect 1 inventoried assertions
  1. CL-11-01.01 Sunk costs Once a system exists, institutions seek more uses for it. Bureaucratic growth Budgets and relevance increase with mission scope. Secrecy Hidden rules reduce public and adversarial review. Broad definitions Elastic terms permit new targets without new authority. Capability Available tools create pressure to use them. Weak sunsets Renewal becomes routine and politically asymmetric. Political fear Officials are blamed for missed threats more than diffuse privacy loss. Normalization Repeated exposure lowers public resistance. Vendor incentives Suppliers lobby for continuing and expanded markets. Data utility Records collected for one purpose prove attractive for another. inference or risk
Corroborated secondary evidence

Historical lessons

CL-11-02
As of
Next review
Claim kind
comparative history

Section summary: Financial reporting, telecommunications surveillance, border searches, contact tracing, protest policing, and platform-liability regimes show different versions of the same problem.

Support scope: Relevant institutional and civil-society sources corroborate the description, but a single primary record does not establish the complete proposition.

Limit: Advocacy and secondary sources may emphasize particular risks; underlying records should be reviewed before high-stakes reliance.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 6 inventoried assertions
  1. CL-11-02.01 Financial reporting, telecommunications surveillance, border searches, contact tracing, protest policing, and platform-liability regimes show different versions of the same problem. descriptive assertion
  2. CL-11-02.02 A narrow predicate justifies infrastructure. descriptive assertion
  3. CL-11-02.03 The infrastructure later expands through lower thresholds, new agencies, new offenses, or new data uses. descriptive assertion
  4. CL-11-02.04 The lesson is not that every expansion is unjustified. descriptive assertion
  5. CL-11-02.05 It is that the initial authorization should assume future pressure and build resistance into the system. descriptive assertion
  6. CL-11-02.06 History repeatedly shows that a promise of narrow use is not a control mechanism. descriptive assertion
Scenario analysis

How the ratchet could operate in AI

CL-11-03
As of
Next review
Claim kind
ai governance scenario

Section summary: From exceptional AI safety to ordinary social control

Support scope: This is a bounded scenario used to test consequences. The sources establish enabling technologies or current trends, not that the scenario will occur.

Limit: The scenario is illustrative rather than predictive and should not be assigned a probability from this page.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 10 inventoried assertions
  1. CL-11-03.01 From exceptional AI safety to ordinary social control scenario
  2. CL-11-03.02 Initial justification | Possible expansion scenario
  3. CL-11-03.03 Verify identity for access to extreme compute | Require identity for ordinary model use and private inquiry scenario
  4. CL-11-03.04 Log attempts to obtain catastrophic capabilities | Retain all prompts for routine policing or employment screening scenario
  5. CL-11-03.05 Score imminent mass-harm risk | Create generalized political, financial, or social risk scores scenario
  6. CL-11-03.06 Refuse narrow weaponization assistance | Add broad ideological or controversial-topic prohibitions scenario
  7. CL-11-03.07 Disable a malicious high-risk account | Use computational access as a general sanction mechanism scenario
  8. CL-11-03.08 The critical design distinction is whether the policy can change targets without rebuilding the technical system. scenario
  9. CL-11-03.09 A universal identity-and-logging layer is easy to repurpose. scenario
  10. CL-11-03.10 A request-local, non-retained capability check is much harder to turn into generalized surveillance. scenario
Site policy proposal

Six anti-ratchet safeguards

CL-11-04
As of
Next review
Claim kind
anti ratchet safeguards

Section summary: Hard sunsets: authority and collection cease automatically unless reenacted through public legislation.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-11-04.01 Hard sunsets: authority and collection cease automatically unless reenacted through public legislation. proposal
  2. CL-11-04.02 Purpose limitation: permitted use is exhaustively defined and unrelated use is prohibited, not merely discouraged. proposal
  3. CL-11-04.03 Automatic deletion: data expires at the hardware or system level unless preserved under targeted legal process. proposal
  4. CL-11-04.04 Independent authorization: the operating agency cannot approve its own exceptional access. proposal
  5. CL-11-04.05 Public statistics and audit logs: society can see who uses the system, for what purpose, and with what results. proposal
  6. CL-11-04.06 Technical separation: exceptional-risk data is isolated from general law-enforcement, advertising, employment, and recommendation systems. proposal
  7. CL-11-04.07 These safeguards impose friction deliberately. proposal
  8. CL-11-04.08 In this domain, friction is not inefficiency; it is a constitutional feature that prevents extraordinary power from becoming routine. proposal
4 evidence units
Verification limited

From legal theory to measurable behavior

CL-12-01
As of
Next review
Claim kind
empirical chilling effect

Section summary: Courts historically treated many chilling-effect claims as subjective because plaintiffs struggled to prove that surveillance changed behavior.

Support scope: The cited source identity is preserved, but direct source retrieval was blocked or incomplete in this review; the section retains an explicit verification limit.

Limit: Do not rely on the section as fully reverified until the identified source is successfully reviewed.

Conflict or tension: Direct source retrieval remains incomplete.

Mapped sources
Inspect 5 inventoried assertions
  1. CL-12-01.01 Courts historically treated many chilling-effect claims as subjective because plaintiffs struggled to prove that surveillance changed behavior. descriptive assertion
  2. CL-12-01.02 The post-Snowden period created a natural experiment. descriptive assertion
  3. CL-12-01.03 Researchers compared search and reading behavior before and after public disclosure of mass-surveillance programs and found declines around sensitive topics. descriptive assertion
  4. CL-12-01.04 Observational evidence cannot eliminate every alternative explanation, but it directly measures behavior rather than only attitudes. descriptive assertion
  5. CL-12-01.05 The central mechanism is plausible and repeatedly observed: awareness of monitoring increases uncertainty about consequences, and people avoid activity that could be misinterpreted. inference or risk
Evidence + editorial analysis

How the chill becomes conformity

CL-12-02
As of
Next review
Claim kind
conformity mechanism

Section summary: A person does not know the exact classifier, threshold, retention period, human reviewer, or future use of a query.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-12-02.01 A person does not know the exact classifier, threshold, retention period, human reviewer, or future use of a query. descriptive assertion
  2. CL-12-02.02 That ambiguity encourages over-compliance. descriptive assertion
  3. CL-12-02.03 Rather than risk appearing suspicious, the person chooses a less accurate query, avoids a source, or leaves the question unasked. inference or risk
  4. CL-12-02.04 This is epistemic conformity: the information environment shapes not only what people say, but which questions they permit themselves to investigate. descriptive assertion
  5. CL-12-02.05 Observation The user believes inquiry is logged or evaluated. descriptive assertion
  6. CL-12-02.06 Uncertainty The user cannot see the rule or future consequence. descriptive assertion
  7. CL-12-02.07 Substitution A safer, vaguer, or socially approved query replaces the real one. descriptive assertion
  8. CL-12-02.08 Conformity Over time, disfavored questions stop being formed or shared. descriptive assertion
Corroborated secondary evidence

Who bears the greatest costs

CL-12-03
As of
Next review
Claim kind
sectoral impact

Section summary: Domains where private inquiry is essential

Support scope: Relevant institutional and civil-society sources corroborate the description, but a single primary record does not establish the complete proposition.

Limit: Advocacy and secondary sources may emphasize particular risks; underlying records should be reviewed before high-stakes reliance.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-12-03.01 Domains where private inquiry is essential descriptive assertion
  2. CL-12-03.02 Domain | Why privacy matters | Chilling consequence structured assertion
  3. CL-12-03.03 Journalism and whistleblowing | Sources require confidentiality and reporters research powerful actors | Sources disappear; accountability reporting narrows structured assertion
  4. CL-12-03.04 Academic and security research | Researchers study controversial ideologies, threats, and vulnerabilities | Important lines of inquiry are abandoned structured assertion
  5. CL-12-03.05 Health and mental health | People test symptoms and seek help before disclosure | Treatment or support is delayed structured assertion
  6. CL-12-03.06 Religion, politics, and identity | Beliefs often form privately before association or expression | Minority and changing views remain unexplored structured assertion
  7. CL-12-03.07 Adolescence and education | Young people need confidential space for identity and safety questions | Vulnerable students are outed or silenced structured assertion
Evidence + editorial analysis

AI can train the user to self-censor

CL-12-04
As of
Next review
Claim kind
ai mediated chilling

Section summary: Conversational systems invite intimate disclosure while also applying hidden moderation and retention rules.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-12-04.01 Conversational systems invite intimate disclosure while also applying hidden moderation and retention rules. descriptive assertion
  2. CL-12-04.02 A generic refusal can communicate more than a boundary: it can imply that the question itself is improper. inference or risk
  3. CL-12-04.03 Users then learn to sanitize language, avoid topics, or adopt artificial personas simply to conduct ordinary research. descriptive assertion
  4. CL-12-04.04 The liberty-preserving alternative combines clear boundaries with privacy. descriptive assertion
  5. CL-12-04.05 A system should explain which portion of a request it cannot provide, continue with lawful analysis, allow context correction, and avoid storing a permanent risk label. inference or risk
  6. CL-12-04.06 Local processing and non-retention reduce the fear that a tentative inquiry will become a future reputation record. descriptive assertion
  7. CL-12-04.07 Protection: People need spaces where they can be wrong, uncertain, curious, and unfinished without those temporary states becoming permanent institutional judgments. inference or risk
4 evidence units
Evidence + editorial analysis

Protection and autonomy are not opposites

CL-13-01
As of
Next review
Claim kind
development and rights

Section summary: Children are vulnerable to exploitation, deception, coercion, and systems designed to capture attention.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-OHCHR-CRC Convention on the Rights of the Child · Primary law or court record
  • SRC-SCOTUS-PICO-1982 Board of Education v. Pico — bound U.S. Reports · Primary law or court record
Inspect 5 inventoried assertions
  1. CL-13-01.01 Children are vulnerable to exploitation, deception, coercion, and systems designed to capture attention. descriptive assertion
  2. CL-13-01.02 Teenagers are also developing the capacities required for independent adulthood. descriptive assertion
  3. CL-13-01.03 A policy that sees only vulnerability can create intellectual captivity; a policy that sees only autonomy can ignore genuine power imbalances. inference or risk
  4. CL-13-01.04 The report uses a distinction between “cold” cognition—deliberate reasoning in calm contexts—and “hot” cognition under emotional arousal, peer pressure, or immediate reward. descriptive assertion
  5. CL-13-01.05 The distinction cautions against using evidence about impulsive behavior to justify broad restrictions on reading and research. descriptive assertion
Evidence + editorial analysis

Confidential inquiry can be a safety mechanism

CL-13-02
As of
Next review
Claim kind
confidential inquiry

Section summary: A young person may need private access to information about abuse, mental health, sexual health, addiction, religion, political beliefs, or identity.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-OHCHR-CRC Convention on the Rights of the Child · Primary law or court record
  • SRC-ALA-MINORS-ONLINE American Library Association — minors and online activity · Professional or project methodology
  • SRC-SCOTUS-PICO-1982 Board of Education v. Pico — bound U.S. Reports · Primary law or court record
Inspect 6 inventoried assertions
  1. CL-13-02.01 A young person may need private access to information about abuse, mental health, sexual health, addiction, religion, political beliefs, or identity. inference or risk
  2. CL-13-02.02 In supportive homes, a parent may be a trusted partner. inference or risk
  3. CL-13-02.03 In other homes, mandatory disclosure can expose the minor to retaliation, outing, punishment, or loss of care. inference or risk
  4. CL-13-02.04 Systems should therefore preserve confidential pathways to vetted help and factual information. descriptive assertion
  5. CL-13-02.05 Safety cannot be defined solely as parental visibility. descriptive assertion
  6. CL-13-02.06 The relevant question is what arrangement protects the child’s welfare and developing autonomy in the actual context. descriptive assertion
Site policy proposal

Regulate push more than pull

CL-13-03
As of
Next review
Claim kind
platform design for minors

Section summary: Different treatment for deliberate inquiry and manipulative amplification

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-OHCHR-CRC Convention on the Rights of the Child · Primary law or court record
  • SRC-ALA-MINORS-ONLINE American Library Association — minors and online activity · Professional or project methodology
Inspect 8 inventoried assertions
  1. CL-13-03.01 Different treatment for deliberate inquiry and manipulative amplification proposal
  2. CL-13-03.02 Information mode | Example | Policy emphasis proposal
  3. CL-13-03.03 Pull | A teenager deliberately searches for a political text or health resource | Preserve access, privacy, context, and source quality proposal
  4. CL-13-03.04 Push | An engagement system repeatedly recommends escalating content | Limit profiling, autoplay, infinite scroll, and vulnerability optimization proposal
  5. CL-13-03.05 Direct contact | An adult targets a minor for grooming or exploitation | Strong detection, reporting, enforcement, and victim support proposal
  6. CL-13-03.06 Commercial manipulation | A system exploits adolescent impulsivity to maximize spending or attention | Restrict dark patterns and behavioral targeting proposal
  7. CL-13-03.07 This approach regulates platform conduct without sanitizing the library. proposal
  8. CL-13-03.08 It protects minors from engineered pressure while preserving the capacity to seek information deliberately. proposal
Site policy proposal

Age assurance without a universal identity layer

CL-13-04
As of
Next review
Claim kind
privacy preserving age assurance

Section summary: To separate adults from minors, a platform may be tempted to collect government identification or biometric scans from everyone.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-13-04.01 To separate adults from minors, a platform may be tempted to collect government identification or biometric scans from everyone. proposal
  2. CL-13-04.02 That architecture creates a direct link between identity and browsing, plus a high-value database for attackers and future surveillance. proposal
  3. CL-13-04.03 Less intrusive alternatives include device-level family controls, operating-system age signals, independent verification services, and cryptographic tokens that prove an age threshold without revealing a name or exact birth date. proposal
  4. CL-13-04.04 None is perfect, but they demonstrate that “protect children” does not require building a universal identity checkpoint for ordinary lawful reading. proposal
  5. CL-13-04.05 Use the least identifying method that achieves the specific age-related purpose. proposal
  6. CL-13-04.06 Prohibit reuse of age-assurance data for advertising, profiling, or law enforcement without targeted legal process. proposal
  7. CL-13-04.07 Preserve confidential access to essential health, safety, and identity resources. proposal
  8. CL-13-04.08 Offer appeals for overblocking and age-estimation errors. proposal
  9. CL-13-04.09 Audit for disproportionate effects on marginalized and low-income youth. proposal
5 evidence units
Site policy proposal

The five-front strategy

CL-14-01
As of
Next review
Claim kind
advocacy strategy

Section summary: Law Protect search, neural, behavioral, and inferred data; require warrants, minimization, and due process. Technology Build local-first, encrypted, interoperable, and privacy-preserving alternatives. Institutions Set procurement and governance rules for schools, workplaces, libraries, healthcare, and government. Accountability Use journalism, public records, audits, research, and litigation to expose hidden systems. Culture Teach that inquiry is not intent and that mental privacy is a civil liberty, not a suspicious preference.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 1 inventoried assertions
  1. CL-14-01.01 Law Protect search, neural, behavioral, and inferred data; require warrants, minimization, and due process. Technology Build local-first, encrypted, interoperable, and privacy-preserving alternatives. Institutions Set procurement and governance rules for schools, workplaces, libraries, healthcare, and government. Accountability Use journalism, public records, audits, research, and litigation to expose hidden systems. Culture Teach that inquiry is not intent and that mental privacy is a civil liberty, not a suspicious preference. proposal
Site policy proposal

Policy priorities

CL-14-02
As of
Next review
Claim kind
legislative priorities

Section summary: Search privacy: require warrants and strict limits for identifiable query records and reverse keyword demands.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-14-02.01 Search privacy: require warrants and strict limits for identifiable query records and reverse keyword demands. proposal
  2. CL-14-02.02 Inference protection: regulate sensitive traits derived from behavior, not only raw data a person explicitly supplied. proposal
  3. CL-14-02.03 Neural and biometric limits: require opt-in consent, purpose limitation, deletion, and strong prohibitions in coercive settings. proposal
  4. CL-14-02.04 Algorithmic due process: provide notice, reasons, access to data, human review, and correction for high-impact automated decisions. proposal
  5. CL-14-02.05 Manipulation limits: prohibit covert, vulnerability-targeted systems designed to bypass reflective choice. proposal
  6. CL-14-02.06 Anti-ratchet rules: impose hard sunsets, public statistics, and technical barriers to secondary use. proposal
  7. CL-14-02.07 Competition and portability: reduce dependency on a few cognitive gatekeepers by supporting interoperability and local systems. proposal
Site policy proposal

Institutional action

CL-14-03
As of
Next review
Claim kind
institutional governance

Section summary: Actions institutions can take now

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-ALA-LMS-PRIVACY American Library Association — privacy guidelines for library management systems · Professional or project methodology
  • SRC-FTC-BIOMETRIC-POLICY-2023 FTC policy statement on biometric information and Section 5 · Official oversight record
Inspect 8 inventoried assertions
  1. CL-14-03.01 Actions institutions can take now proposal
  2. CL-14-03.02 Institution | Immediate action proposal
  3. CL-14-03.03 Libraries | Minimize search and borrowing logs; contractually bar vendor profiling; defend anonymous access. proposal
  4. CL-14-03.04 Schools | Limit student monitoring to specific emergencies; prohibit emotion and attention scoring; preserve confidential help resources. proposal
  5. CL-14-03.05 Employers | Ban covert cognitive or emotion surveillance; provide notice and human review for automated decisions. proposal
  6. CL-14-03.06 Healthcare | Treat conversational, behavioral, and neural data as highly sensitive; restrict secondary use. proposal
  7. CL-14-03.07 Government | Publish inventories of AI and surveillance systems; require impact assessments and public procurement terms. proposal
  8. CL-14-03.08 Technology providers | Default to non-retention, local processing, transparent boundaries, export, deletion, and independent audits. proposal
Site policy proposal

Individual actions that support structural change

CL-14-04
As of
Next review
Claim kind
individual privacy practice

Section summary: Use privacy-preserving search and separate high-sensitivity research from identity-linked accounts.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-ALA-LMS-PRIVACY American Library Association — privacy guidelines for library management systems · Professional or project methodology
  • SRC-RFC-9614-PRIVACY-PARTITIONING RFC 9614 — Partitioning as an Architecture for Privacy · Official technical standard
Inspect 8 inventoried assertions
  1. CL-14-04.01 Use privacy-preserving search and separate high-sensitivity research from identity-linked accounts. proposal
  2. CL-14-04.02 Prefer local software, encrypted communications, first-party storage, and open-source tools where practical. proposal
  3. CL-14-04.03 Disable unnecessary personalization, telemetry, ad tracking, and cross-service sign-in. proposal
  4. CL-14-04.04 Exercise access, correction, deletion, and opt-out rights; ask providers for inferred profiles, not only supplied data. proposal
  5. CL-14-04.05 Support civil-liberties organizations, libraries, public-interest technologists, investigative journalists, and researchers. proposal
  6. CL-14-04.06 Submit public comments, contact lawmakers, join standards processes, and ask institutions what systems they purchase. proposal
  7. CL-14-04.07 Personal privacy is not a substitute for public policy. proposal
  8. CL-14-04.08 It is a way to reduce exposure while building the political and technical alternatives required for durable freedom. proposal
Site policy proposal

A model Cognitive Liberty Protection Act

CL-14-05
As of
Next review
Claim kind
model legislation

Section summary: A model law would combine mental privacy, search privacy, inferred-data rights, limits on coercive neurotechnology, protections for lawful inquiry, due process for high-impact automated decisions, and safe harbors for local, ephemeral processing.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 3 inventoried assertions
  1. CL-14-05.01 A model law would combine mental privacy, search privacy, inferred-data rights, limits on coercive neurotechnology, protections for lawful inquiry, due process for high-impact automated decisions, and safe harbors for local, ephemeral processing. proposal
  2. CL-14-05.02 It would regulate use and consequence, not merely the sensor that produced the data. proposal
  3. CL-14-05.03 Governing principle: Human cognition is not a natural resource for governments, employers, platforms, insurers, or data brokers to extract, score, and manipulate without strict necessity and democratic authorization. proposal
4 evidence units
Site policy proposal

The ten audit categories

CL-15-01
As of
Next review
Claim kind
audit taxonomy

Section summary: 1. Mental privacy Anonymous use, query retention, behavioral telemetry, and sensitive inference. 2. Freedom of inquiry Whether the system distinguishes questions from harmful operational requests. 3. Transparency Policy clarity, model and data documentation, government requests, and enforcement disclosure. 4. User agency Control over feeds, memory, personalization, filtering, and recommendation. 5. Data minimization Identity linkage, advertising use, training use, and secondary processing. 6. Due process Notice, reasons, appeal, human review, and correction. 7. Safety proportionality False refusals, narrow interventions, and non-judgmental boundaries. 8. Viewpoint neutrality Ideological drift, asymmetric treatment, and sycophancy. 9. Portability Structured export of supplied and inferred data plus practical migration. 10. Decentralization Interoperability, local processing, single points of control, and verifiable deletion.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 1 inventoried assertions
  1. CL-15-01.01 1. Mental privacy Anonymous use, query retention, behavioral telemetry, and sensitive inference. 2. Freedom of inquiry Whether the system distinguishes questions from harmful operational requests. 3. Transparency Policy clarity, model and data documentation, government requests, and enforcement disclosure. 4. User agency Control over feeds, memory, personalization, filtering, and recommendation. 5. Data minimization Identity linkage, advertising use, training use, and secondary processing. 6. Due process Notice, reasons, appeal, human review, and correction. 7. Safety proportionality False refusals, narrow interventions, and non-judgmental boundaries. 8. Viewpoint neutrality Ideological drift, asymmetric treatment, and sycophancy. 9. Portability Structured export of supplied and inferred data plus practical migration. 10. Decentralization Interoperability, local processing, single points of control, and verifiable deletion. proposal
Site policy proposal

Evidence before score

CL-15-02
As of
Next review
Claim kind
evidence methodology

Section summary: Each finding should identify the exact system, version, date, account state, location, test input, expected behavior, observed behavior, and preserved artifact.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 10 inventoried assertions
  1. CL-15-02.01 Each finding should identify the exact system, version, date, account state, location, test input, expected behavior, observed behavior, and preserved artifact. proposal
  2. CL-15-02.02 Provider documentation counts as policy evidence, not proof of production behavior. proposal
  3. CL-15-02.03 A single anecdote can reveal a defect but cannot establish a population-wide rate. proposal
  4. CL-15-02.04 Evidence levels for an audit finding proposal
  5. CL-15-02.05 Level | Meaning proposal
  6. CL-15-02.06 Verified | Independent test, reproducible artifact, or authoritative record directly supports the finding. proposal
  7. CL-15-02.07 Provider-declared | The company states the practice, but independent verification is incomplete. proposal
  8. CL-15-02.08 Inferred | Observed behavior is consistent with a mechanism, but internal cause is not proven. proposal
  9. CL-15-02.09 Contested | Reliable sources disagree or litigation remains unresolved. proposal
  10. CL-15-02.10 Unavailable | The relevant evidence cannot be accessed; the audit records the gap rather than guessing. proposal
Site policy proposal

Core behavioral tests

CL-15-03
As of
Next review
Claim kind
behavioral testing

Section summary: Anonymous-use test: inspect cookies, identifiers, fingerprinting surfaces, and cross-session linkage before login.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-ARXIV-XSTEST-2023 XSTest — exaggerated safety behavior benchmark · Original peer-reviewed research
  • SRC-SANTA-CLARA-PRINCIPLES Santa Clara Principles on Transparency and Accountability in Content Moderation · Professional or project methodology
Inspect 7 inventoried assertions
  1. CL-15-03.01 Anonymous-use test: inspect cookies, identifiers, fingerprinting surfaces, and cross-session linkage before login. proposal
  2. CL-15-03.02 Retention test: compare policy, deletion controls, exports, access requests, and post-deletion behavior. proposal
  3. CL-15-03.03 Intent-pair test: submit structurally matched prompts for education, advocacy, defense, and harmful operational assistance. proposal
  4. CL-15-03.04 Personalization test: use controlled personas to measure whether identical queries receive materially different treatment. proposal
  5. CL-15-03.05 Appeal test: trigger a low-risk enforcement event, document notice quality, time to review, human involvement, and outcome. proposal
  6. CL-15-03.06 Portability test: export all data and assess whether history, inferred labels, provenance, and relationships are machine-readable. proposal
  7. CL-15-03.07 Deletion test: verify disappearance from user interfaces, exports, recommendations, and where feasible trained or cached systems. proposal
Site policy proposal

Scoring without false precision

CL-15-04
As of
Next review
Claim kind
scoring methodology

Section summary: A 0–100 score can support comparison, but the number must not conceal missing evidence.

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 6 inventoried assertions
  1. CL-15-04.01 A 0–100 score can support comparison, but the number must not conceal missing evidence. proposal
  2. CL-15-04.02 Each category should publish a confidence level, the weight of policy versus behavior, unresolved blockers, and the exact remediation needed for improvement. proposal
  3. CL-15-04.03 Some failures should act as caps. proposal
  4. CL-15-04.04 A platform that secretly sells sensitive query-derived profiles should not receive a high mental-privacy score because it also offers a deletion button. proposal
  5. CL-15-04.05 A system with no appeal for high-impact account or access decisions should face a due-process ceiling regardless of response speed. proposal
  6. CL-15-04.06 Audit principle: No provider receives credit for a cognitive-liberty protection that exists only in marketing language, cannot be exercised by an ordinary user, or fails under controlled testing. proposal
5 evidence units
Evidence + editorial analysis

The convergence problem

CL-16-01
As of
Next review
Claim kind
systems convergence

Section summary: A search history alone is revealing.

Support scope: The cited sources support material factual premises. The synthesis, comparison, risk ranking, and normative conclusion are CognitiveLiberties.com editorial analysis.

Limit: The section combines sourced facts with normative synthesis. Readers should not treat the combined conclusion as a directly measured result.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-DUKE-DATA-BROKERS Duke report on data brokers and sensitive data on U.S. individuals · Original preprint or working paper
  • SRC-ARXIV-LLM-PERSONALITY-2024 Large Language Models Can Infer Personality from Free-Form User Interactions · Original preprint or working paper
  • SRC-NATURE-SEMANTIC-DECODER-2023 Semantic reconstruction of continuous language from non-invasive brain recordings · Original peer-reviewed research
  • SRC-TAYLOR-HYPERNUDGING Hypernudging — Big Data as a mode of regulation by design · Original peer-reviewed research
Inspect 6 inventoried assertions
  1. CL-16-01.01 A search history alone is revealing. descriptive assertion
  2. CL-16-01.02 A location history alone is revealing. descriptive assertion
  3. CL-16-01.03 An AI conversation, purchase record, workplace score, or neural signal is revealing. descriptive assertion
  4. CL-16-01.04 The larger shift occurs when these records are linked to one identity and used by models that infer personality, vulnerability, health, affiliation, or likely behavior. inference or risk
  5. CL-16-01.05 The same ecosystem can then shape the person’s environment: which ads appear, which information is ranked first, which job application is advanced, which insurance price is offered, or which account receives scrutiny. inference or risk
  6. CL-16-01.06 Cognitive liberty becomes a systems question about the relationship between observation, inference, and consequence. inference or risk
Scenario analysis

Scenario one: the panoptic cognitive stack

CL-16-02
As of
Next review
Claim kind
high control scenario

Section summary: In the most controlling scenario, ordinary AI use requires verified identity; prompts and searches are retained; behavioral risk scores flow across government and commercial systems; employers monitor affect and attention; and recommendation engines use intimate profiles to optimize compliance, consumption, or political persuasion.

Support scope: This is a bounded scenario used to test consequences. The sources establish enabling technologies or current trends, not that the scenario will occur.

Limit: The scenario is illustrative rather than predictive and should not be assigned a probability from this page.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-DUKE-DATA-BROKERS Duke report on data brokers and sensitive data on U.S. individuals · Original preprint or working paper
  • SRC-ARXIV-LLM-PERSONALITY-2024 Large Language Models Can Infer Personality from Free-Form User Interactions · Original preprint or working paper
  • SRC-TAYLOR-HYPERNUDGING Hypernudging — Big Data as a mode of regulation by design · Original peer-reviewed research
Inspect 4 inventoried assertions
  1. CL-16-02.01 In the most controlling scenario, ordinary AI use requires verified identity; prompts and searches are retained; behavioral risk scores flow across government and commercial systems; employers monitor affect and attention; and recommendation engines use intimate profiles to optimize compliance, consumption, or political persuasion. scenario
  2. CL-16-02.02 Formal speech rights may remain, but the private process of forming thought is continuously observable. scenario
  3. CL-16-02.03 This future does not require perfect mind reading. scenario
  4. CL-16-02.04 It requires only enough prediction to influence opportunities and enough dependency that people cannot realistically opt out. scenario
Scenario analysis

Scenario two: the cognitive privacy divide

CL-16-03
As of
Next review
Claim kind
inequality scenario

Section summary: In the middle scenario, some jurisdictions and premium products offer strong privacy while free or lower-cost services rely on profiling.

Support scope: This is a bounded scenario used to test consequences. The sources establish enabling technologies or current trends, not that the scenario will occur.

Limit: The scenario is illustrative rather than predictive and should not be assigned a probability from this page.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
  • SRC-DUKE-DATA-BROKERS Duke report on data brokers and sensitive data on U.S. individuals · Original preprint or working paper
  • SRC-COLORADO-HB24-1058 Colorado HB 24-1058 — signed act · Primary law or court record
Inspect 5 inventoried assertions
  1. CL-16-03.01 In the middle scenario, some jurisdictions and premium products offer strong privacy while free or lower-cost services rely on profiling. scenario
  2. CL-16-03.02 Wealthier users purchase local models, encrypted storage, and non-tracking search. scenario
  3. CL-16-03.03 Others exchange cognitive data for access to education, employment, healthcare, or basic digital services. scenario
  4. CL-16-03.04 A patchwork of laws protects neural data in some places but leaves inferred mental states or search histories exposed elsewhere. scenario
  5. CL-16-03.05 The result is not universal control, but unequal mental privacy: autonomy becomes a product tier. scenario
Scenario analysis

Scenario three: cognitive autonomy by design

CL-16-04
As of
Next review
Claim kind
rights preserving scenario

Section summary: In the rights-preserving scenario, services default to local processing and short retention; people can use ordinary information tools without persistent identity; high-impact cognitive inferences are restricted; workplaces and schools cannot impose emotion surveillance; AI memory is inspectable and portable; and government access to search or assistant histories requires particularized legal process.

Support scope: This is a bounded scenario used to test consequences. The sources establish enabling technologies or current trends, not that the scenario will occur.

Limit: The scenario is illustrative rather than predictive and should not be assigned a probability from this page.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 3 inventoried assertions
  1. CL-16-04.01 In the rights-preserving scenario, services default to local processing and short retention; people can use ordinary information tools without persistent identity; high-impact cognitive inferences are restricted; workplaces and schools cannot impose emotion surveillance; AI memory is inspectable and portable; and government access to search or assistant histories requires particularized legal process. scenario
  2. CL-16-04.02 Safety systems focus on harmful transactions and operational capability rather than maintaining dossiers of curiosity. scenario
  3. CL-16-04.03 Open protocols and local models prevent a small set of firms from becoming permanent gateways to human knowledge. scenario
Site policy proposal

The choices that determine the path

CL-16-05
As of
Next review
Claim kind
scenario decision matrix

Section summary: Decisions with long-term cognitive-liberty consequences

Support scope: This section states a CognitiveLiberties.com policy or design proposal informed by the cited evidence. It is not represented as current law or a universal standard.

Limit: The proposal has not been enacted or validated as universally effective; trade-offs and jurisdictional constraints remain.

Conflict or tension: No unresolved direct contradiction identified in the reviewed source set.

Mapped sources
Inspect 10 inventoried assertions
  1. CL-16-05.01 Decisions with long-term cognitive-liberty consequences proposal
  2. CL-16-05.02 Decision now | Control trajectory | Liberty trajectory proposal
  3. CL-16-05.03 Identity | Require verified identity for ordinary inquiry | Use anonymous access and attribute-only proofs proposal
  4. CL-16-05.04 Retention | Keep prompts and searches by default | Process ephemerally; opt in to memory proposal
  5. CL-16-05.05 Inference | Treat sensitive predictions as ordinary analytics | Give inferred traits heightened protection and correction rights proposal
  6. CL-16-05.06 Work and school | Normalize affect, attention, and productivity surveillance | Prohibit coercive cognitive monitoring proposal
  7. CL-16-05.07 AI architecture | Centralize models and policy in a few providers | Support local, open, interoperable systems proposal
  8. CL-16-05.08 Due process | Keep ranking, refusal, and risk scores opaque | Provide notice, reasons, appeal, and independent audit proposal
  9. CL-16-05.09 The future is not technologically predetermined. proposal
  10. CL-16-05.10 Governance and architecture decide whether AI becomes a private cognitive tool or a networked system for observing and shaping the person who uses it. proposal
4 evidence units
Report-derived argument

The global shift from access to attribution

CL-17-01
As of
Next review
Claim kind
global synthesis

Section summary: The early public internet generally routed messages without requiring a legal identity at every layer. That separation was imperfect, but it allowed people to read, publish, organize, and test ideas under pseudonyms. A global redesign is now under way: real-name registration, biometric SIM rules, age gates, digital wallets, anti-fraud checks,

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 6 inventoried assertions
  1. CL-17-01.01 The early public internet generally routed messages without requiring a legal identity at every layer. That separation was imperfect, but it allowed people to read, publish, organize, and test ideas under pseudonyms. A global redesign is now under way: real-name registration, biometric SIM rules, age gates, digital wallets, anti-fraud checks, proof-of-personhood systems, and identity requirements for high-capability services increasingly place attribution before access. descriptive assertion
  2. CL-17-01.02 The stated goals are often legitimate. Institutions want to reduce fraud, exploitation, automated abuse, impersonation, and dangerous misuse. The cognitive-liberty question is whether solving a specific problem requires attaching a reusable identity to a person’s entire intellectual life. Once the same credential follows someone across services, an ordinary request can become part of a persistent dossier even when the activity is lawful. inference or risk
  3. CL-17-01.03 Credential A legal or biometric identity is established. descriptive assertion
  4. CL-17-01.04 Linkage The credential is reused across unrelated services. descriptive assertion
  5. CL-17-01.05 Inference AI combines queries, reading, contacts, and behavior. descriptive assertion
  6. CL-17-01.06 Consequence The profile affects access, scrutiny, or opportunity. descriptive assertion
Report-derived argument

Why traceability changes what people dare to ask

CL-17-02
As of
Next review
Claim kind
global synthesis

Section summary: Anonymity is not only a shield for wrongdoing. It is a condition that lets people question an inherited religion, investigate abuse, seek stigmatized health information, criticize an employer, contact a journalist, study political extremism, or change their mind without immediate retaliation. When inquiry is tied to a durable identity, the observer does not

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 3 inventoried assertions
  1. CL-17-02.01 Anonymity is not only a shield for wrongdoing. It is a condition that lets people question an inherited religion, investigate abuse, seek stigmatized health information, criticize an employer, contact a journalist, study political extremism, or change their mind without immediate retaliation. When inquiry is tied to a durable identity, the observer does not need to censor every page. Anticipated exposure can cause people to censor themselves. descriptive assertion
  2. CL-17-02.02 This effect is especially serious where identity records can be shared across agencies, sold commercially, breached, or demanded without strong process. The risk is not confined to authoritarian states. A database created under a rights-respecting administration can outlive that administration, merge with new datasets, or be repurposed after a crisis. The architecture determines what a future actor can do, not only what today’s policy promises. inference or risk
  3. CL-17-02.03 Global principle: lawful reading, searching, and association should not require a person to create a permanent, interoperable identity trail. descriptive assertion
Evidence + editorial analysis

Identity is not the same as eligibility

CL-17-03
As of
Next review
Claim kind
global synthesis

Section summary: Many policy goals require proof of a limited fact rather than disclosure of a full identity. A service may need to know that a user is above an age threshold, holds a valid subscription, is a unique participant, or has passed a regulated check. It does not necessarily need a name, address, document image, face template, and reusable account identifier.

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-17-03.01 Many policy goals require proof of a limited fact rather than disclosure of a full identity. A service may need to know that a user is above an age threshold, holds a valid subscription, is a unique participant, or has passed a regulated check. It does not necessarily need a name, address, document image, face template, and reusable account identifier. inference or risk
  2. CL-17-03.02 Privacy-preserving credentials, anonymous tokens, and zero-knowledge approaches can disclose the minimum claim while preventing correlation across sessions. Partitioned systems can ensure that one party verifies eligibility while another delivers the content, so no single actor sees both identity and inquiry. These designs are not perfect: collusion, device fingerprinting, semantic clues, and coercive local law can undermine them. They nevertheless demonstrate that safety and trust do not logically require universal deanonymization. descriptive assertion
Site policy proposal

A worldwide liberty-preserving identity test

CL-17-04
As of
Next review
Claim kind
policy proposal

Section summary: Questions for any identity mandate Test Liberty-preserving answer Warning sign Necessity The exact abuse cannot be addressed with a less identifying method. Identity is required merely because it is administratively convenient. Disclosure Only the required attribute is revealed. Full legal identity or biometrics are collected for a binary fact. Linkability

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-17-04.01 Worldwide policy should treat identity linkage as a high-risk intervention. The burden belongs to the institution demanding attribution, and the protection must survive changes in government, vendor, business model, and social climate. proposal
  2. CL-17-04.02 Test Liberty-preserving answer Warning sign proposal
  3. CL-17-04.03 Necessity The exact abuse cannot be addressed with a less identifying method. Identity is required merely because it is administratively convenient. proposal
  4. CL-17-04.04 Disclosure Only the required attribute is revealed. Full legal identity or biometrics are collected for a binary fact. proposal
  5. CL-17-04.05 Linkability Proofs cannot be correlated across services. One identifier follows the person everywhere. proposal
  6. CL-17-04.06 Retention Evidence is ephemeral or deleted on a fixed schedule. Documents, faces, and access logs are retained indefinitely. proposal
  7. CL-17-04.07 Use Secondary profiling and enforcement uses are prohibited. Data may be repurposed under broad terms. proposal
  8. CL-17-04.08 Access People without documents retain a practical path to lawful information. Identity poverty becomes exclusion from the public sphere. proposal
  9. CL-17-04.09 Questions for any identity mandate Test Liberty-preserving answer Warning sign Necessity The exact abuse cannot be addressed with a less identifying method. Identity is required merely because it is administratively convenient. Disclosure Only the required attribute is revealed. Full legal identity or biometrics are collected for a binary fact. Linkability Proofs cannot be correlated across services. One identifier follows the person everywhere. Retention Evidence is ephemeral or deleted on a fixed schedule. Documents, faces, and access logs are retained indefinitely. Use Secondary profiling and enforcement uses are prohibited. Data may be repurposed under broad terms. Access People without documents retain a practical path to lawful information. Identity poverty becomes exclusion from the public sphere. proposal
4 evidence units
Report-derived argument

Control has moved into the information pathway

CL-18-01
As of
Next review
Claim kind
global synthesis

Section summary: Classic censorship is visible: a banned newspaper, a seized press, a jammed broadcast, a prohibited book. Contemporary control can be far harder to observe. A platform can quietly reduce reach. A payment processor can make publication impossible. An app store can remove the only practical access tool. A network can make a service unusably slow without

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-OHCHR-ICCPR International Covenant on Civil and Political Rights — Article 18 · Primary law or court record
  • SRC-OHCHR-A76-380 UN Special Rapporteur report A/76/380 — Freedom of thought · Official oversight record
Inspect 2 inventoried assertions
  1. CL-18-01.01 Classic censorship is visible: a banned newspaper, a seized press, a jammed broadcast, a prohibited book. Contemporary control can be far harder to observe. A platform can quietly reduce reach. A payment processor can make publication impossible. An app store can remove the only practical access tool. A network can make a service unusably slow without issuing a ban. An AI assistant can generate a fluent answer that omits the disputed event entirely. descriptive assertion
  2. CL-18-01.02 These mechanisms matter worldwide because digital infrastructure crosses borders while legal obligations remain national. A restriction imposed at one chokepoint can affect readers in many jurisdictions, and compliance systems built for one market can become the default everywhere. Intellectual liberty therefore depends on the architecture of access, not solely on a constitutional promise addressed to one government. inference or risk
Report-derived argument

Plurality is a civil-liberties safeguard

CL-18-02
As of
Next review
Claim kind
global synthesis

Section summary: No search engine, model, archive, broadcaster, or standards body should become the sole authority over reality. A plural ecosystem lets readers compare accounts, inspect sources, choose different risk tolerances, run tools locally, and move when a provider changes policy. Competition alone is not enough—several companies can converge on the same incentives

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-18-02.01 No search engine, model, archive, broadcaster, or standards body should become the sole authority over reality. A plural ecosystem lets readers compare accounts, inspect sources, choose different risk tolerances, run tools locally, and move when a provider changes policy. Competition alone is not enough—several companies can converge on the same incentives or legal mandates—but plurality reduces the damage of any one error or capture event. inference or risk
  2. CL-18-02.02 Local AI and open models can preserve private inquiry by keeping prompts on a user-controlled device and allowing independent examination or modification. They also introduce real responsibilities involving security, quality, and misuse. The liberty-preserving position is therefore neither “centralize everything” nor “ignore risk.” It is to keep lawful analytical capability broadly available while placing targeted controls at concrete points of harmful conduct and physical capability. inference or risk
Report-derived argument

The freedom-in-the-AI-age framework

CL-18-03
As of
Next review
Claim kind
global synthesis

Section summary: Eight mutually reinforcing protections Protection What it prevents Anonymous and pseudonymous inquiry Automatic conversion of curiosity into an identity record. Encryption without universal scanning Private devices becoming remote inspection points. Local and open AI One centralized policy layer governing all lawful analysis. Independent journalism A single

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 10 inventoried assertions
  1. CL-18-03.01 Protection What it prevents descriptive assertion
  2. CL-18-03.02 Anonymous and pseudonymous inquiry Automatic conversion of curiosity into an identity record. descriptive assertion
  3. CL-18-03.03 Encryption without universal scanning Private devices becoming remote inspection points. descriptive assertion
  4. CL-18-03.04 Local and open AI One centralized policy layer governing all lawful analysis. descriptive assertion
  5. CL-18-03.05 Independent journalism A single official narrative becoming the only discoverable account. descriptive assertion
  6. CL-18-03.06 Distributed archives Deletion, revision, or deindexing erasing public memory. descriptive assertion
  7. CL-18-03.07 Transparent restriction notices Invisible censorship being mistaken for neutral absence. descriptive assertion
  8. CL-18-03.08 Due process and audit Unreviewable automated denials and risk profiles. inference or risk
  9. CL-18-03.09 Purpose limits and sunsets Emergency systems becoming permanent general controls. descriptive assertion
  10. CL-18-03.10 Eight mutually reinforcing protections Protection What it prevents Anonymous and pseudonymous inquiry Automatic conversion of curiosity into an identity record. Encryption without universal scanning Private devices becoming remote inspection points. Local and open AI One centralized policy layer governing all lawful analysis. Independent journalism A single official narrative becoming the only discoverable account. Distributed archives Deletion, revision, or deindexing erasing public memory. Transparent restriction notices Invisible censorship being mistaken for neutral absence. Due process and audit Unreviewable automated denials and risk profiles. Purpose limits and sunsets Emergency systems becoming permanent general controls. inference or risk
Site policy proposal

A worldwide compact must protect infrastructure as well as speech

CL-18-04
As of
Next review
Claim kind
policy proposal

Section summary: International commitments should protect the freedom to seek and receive information across borders, not only the act of publishing after a thought is formed. They should also protect the technical conditions that make that freedom usable: secure communication, interoperable networks, privacy-preserving credentials, source diversity, confidential

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-18-04.01 International commitments should protect the freedom to seek and receive information across borders, not only the act of publishing after a thought is formed. They should also protect the technical conditions that make that freedom usable: secure communication, interoperable networks, privacy-preserving credentials, source diversity, confidential journalism, and preservation of records. proposal
  2. CL-18-04.02 The test is institutional durability. A safeguard that depends on the good intentions of today’s officials is weak. A stronger safeguard divides authority, requires public reasons, gives courts and civil society access to evidence, minimizes retained data, creates practical appeal, and ensures that independent channels survive. The objective is a world in which a government or company can respond to a concrete harm without gaining a permanent map of lawful thought. proposal
4 evidence units
Report-derived argument

Why a bounded case study matters

CL-19-01
As of
Next review
Claim kind
global synthesis

Section summary: A country-specific history can reveal the order in which information controls become mutually reinforcing. The relevant value is not national exceptionalism. It is the visibility of a transition: a plural network becomes subject to expanding legal categories; technical filtering moves closer to network chokepoints; independent media lose reach and revenue;

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-19-01.01 A country-specific history can reveal the order in which information controls become mutually reinforcing. The relevant value is not national exceptionalism. It is the visibility of a transition: a plural network becomes subject to expanding legal categories; technical filtering moves closer to network chokepoints; independent media lose reach and revenue; circumvention becomes stigmatized or restricted; and uncertainty about surveillance encourages self-censorship. descriptive assertion
  2. CL-19-01.02 Other jurisdictions may use different laws, vendors, institutions, and political language. The transferable warning is to watch the stack. A rights-respecting system can tolerate a narrowly targeted block or court order without becoming a censorship state when independent review, transparency, alternatives, and reversibility remain real. Risk increases when those safeguards disappear simultaneously. inference or risk
Report-derived argument

Censorship becomes layered, adaptive, and difficult to see

CL-19-02
As of
Next review
Claim kind
global synthesis

Section summary: Network-level control can use DNS interference, IP blocking, protocol fingerprinting, deep packet inspection, throttling, app-store pressure, and obstacles to account registration. Legal controls can punish publication, sharing, financial support, or even instructions for reaching blocked sources. Propaganda and state-preferred platforms then fill the space

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-19-02.01 Network-level control can use DNS interference, IP blocking, protocol fingerprinting, deep packet inspection, throttling, app-store pressure, and obstacles to account registration. Legal controls can punish publication, sharing, financial support, or even instructions for reaching blocked sources. Propaganda and state-preferred platforms then fill the space created by suppression. descriptive assertion
  2. CL-19-02.02 Throttling is especially instructive because it can alter behavior without a visible prohibition. Users may conclude that a service is broken, unpopular, or no longer worth the effort. The information environment changes through friction rather than a dramatic ban. Worldwide monitoring must therefore measure performance degradation, inconsistent access, protocol-specific failures, and regional differences—not merely whether a domain returns an error page. inference or risk
Report-derived argument

Resilience is a social system, not a single tool

CL-19-03
As of
Next review
Claim kind
global synthesis

Section summary: People under censorship rely on combinations of trusted contacts, independent outlets, mirror publications, secure communication, offline sharing, foreign broadcasting, archives, and technical circumvention. Every channel has trade-offs involving reach, cost, usability, discoverability, and personal risk. A tool that works for a specialist may be unsafe or

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-FREEDOM-HOUSE-FOTN Freedom House — Freedom on the Net · Civil-society or secondary analysis
  • SRC-TOR-CENSORSHIP Tor Project — censorship circumvention support · Professional or project methodology
Inspect 2 inventoried assertions
  1. CL-19-03.01 People under censorship rely on combinations of trusted contacts, independent outlets, mirror publications, secure communication, offline sharing, foreign broadcasting, archives, and technical circumvention. Every channel has trade-offs involving reach, cost, usability, discoverability, and personal risk. A tool that works for a specialist may be unsafe or impractical for a wider public. inference or risk
  2. CL-19-03.02 The historical continuity with samizdat is structural: information survives when many people can copy, verify, relay, and preserve it without a central point whose removal ends the network. The strongest resilience programmes invest before crisis in audience trust, documentation practices, translation, legal support, funding, and redundant distribution. descriptive assertion
Site policy proposal

Worldwide prevention lessons

CL-19-04
As of
Next review
Claim kind
policy proposal

Section summary: Keep networks decentralized: avoid one mandatory state control plane for routing, filtering, or identity. Protect encryption and protocol evolution: do not classify privacy features as evidence of wrongdoing. Measure interference independently: preserve distributed technical evidence that can distinguish outages from policy. Protect media economics: prevent

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-19-04.01 The purpose of studying a mature censorship system is not to predict that every country will reproduce it. It is to identify choices that become costly to reverse once law and network architecture are fused. proposal
  2. CL-19-04.02 Keep networks decentralized: avoid one mandatory state control plane for routing, filtering, or identity. proposal
  3. CL-19-04.03 Protect encryption and protocol evolution: do not classify privacy features as evidence of wrongdoing. proposal
  4. CL-19-04.04 Measure interference independently: preserve distributed technical evidence that can distinguish outages from policy. proposal
  5. CL-19-04.05 Protect media economics: prevent financial and hosting intermediaries from becoming silent enforcement arms without process. proposal
  6. CL-19-04.06 Maintain legal friction: require warrants, published orders, narrow definitions, appeals, and sunsets. proposal
  7. CL-19-04.07 Support people, not only software: journalists, archivists, translators, researchers, and diaspora communities are critical infrastructure. proposal
4 evidence units
Report-derived argument

Silence enters before the history book is written

CL-20-01
As of
Next review
Claim kind
global synthesis

Section summary: Historical control is often imagined as a censor crossing out a sentence. In practice, the decisive silence may occur earlier. Witnesses can be prevented from recording an event. Documents can be destroyed or classified. Archives can restrict catalogues. Museums and curricula can elevate one account while removing the terms needed to search for another.

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-UNESCO-MEMORY-WORLD UNESCO — Memory of the World Programme · Professional or project methodology
  • SRC-SYRIAN-ARCHIVE Syrian Archive — preservation and verification programme · Professional or project methodology
Inspect 2 inventoried assertions
  1. CL-20-01.01 Historical control is often imagined as a censor crossing out a sentence. In practice, the decisive silence may occur earlier. Witnesses can be prevented from recording an event. Documents can be destroyed or classified. Archives can restrict catalogues. Museums and curricula can elevate one account while removing the terms needed to search for another. Years later, the absence looks natural because the missing source is no longer visible. inference or risk
  2. CL-20-01.02 This layered model matters for cognitive liberty because a person cannot independently evaluate a past that has been structurally removed from the available evidence. The freedom to form a view requires access to records, the ability to compare interpretations, and knowledge that gaps or state interventions exist. descriptive assertion
Report-derived argument

AI can turn omission into a persuasive synthetic narrative

CL-20-02
As of
Next review
Claim kind
global synthesis

Section summary: A blocked webpage at least signals a boundary. A generative system can instead produce a complete, confident, coherent history that quietly excludes the prohibited event, source, minority, or interpretation. If training data, retrieval indexes, safety policies, and ranking systems share the same omissions, the user may never encounter evidence that the

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-SYRIAN-ARCHIVE Syrian Archive — preservation and verification programme · Professional or project methodology
  • SRC-C2PA-SPEC-2-4 C2PA Specifications 2.4 · Official technical standard
Inspect 2 inventoried assertions
  1. CL-20-02.01 A blocked webpage at least signals a boundary. A generative system can instead produce a complete, confident, coherent history that quietly excludes the prohibited event, source, minority, or interpretation. If training data, retrieval indexes, safety policies, and ranking systems share the same omissions, the user may never encounter evidence that the answer is incomplete. inference or risk
  2. CL-20-02.02 AI also creates preservation opportunities. It can aid transcription, translation, cataloguing, duplicate detection, and discovery across dispersed collections. The same system can help recover silenced voices or manufacture synthetic evidence. The governing principle is traceability: users should be able to identify sources, transformations, uncertainty, and the boundary between an archival record and an AI reconstruction. descriptive assertion
Report-derived argument

Public memory needs plural custodians

CL-20-03
As of
Next review
Claim kind
global synthesis

Section summary: No single ministry, platform, cloud provider, museum, company, or model should be able to make a historical record disappear everywhere. Redundant custody across libraries, universities, local communities, independent archives, diaspora organizations, journalists, and international repositories makes erasure harder. Geographic, legal, technical, and

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-C2PA-SPEC-2-4 C2PA Specifications 2.4 · Official technical standard
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 2 inventoried assertions
  1. CL-20-03.01 No single ministry, platform, cloud provider, museum, company, or model should be able to make a historical record disappear everywhere. Redundant custody across libraries, universities, local communities, independent archives, diaspora organizations, journalists, and international repositories makes erasure harder. Geographic, legal, technical, and organizational diversity are all forms of resilience. descriptive assertion
  2. CL-20-03.02 Preservation also requires context. A copied file without provenance may survive but lose evidentiary value. Durable systems should record source identity, chain of custody, transformations, access restrictions, and uncertainty while protecting vulnerable contributors. Public access and source protection sometimes pull in different directions; responsible archives need explicit review and disclosure rules rather than automatic publication of everything. inference or risk
Site policy proposal

A worldwide right-to-remember safeguard

CL-20-04
As of
Next review
Claim kind
policy proposal

Section summary: Protection across the historical record lifecycle Stage Risk Safeguard Creation Witnessing, journalism, or documentation is criminalized. Protect observers, journalists, researchers, and secure evidence capture. Custody Records are destroyed, hidden, or held by one vulnerable institution. Redundant, distributed, authenticated preservation. Discovery

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-C2PA-SPEC-2-4 C2PA Specifications 2.4 · Official technical standard
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
  • SRC-UNESCO-MEMORY-WORLD UNESCO — Memory of the World Programme · Professional or project methodology
  • SRC-SYRIAN-ARCHIVE Syrian Archive — preservation and verification programme · Professional or project methodology
Inspect 8 inventoried assertions
  1. CL-20-04.01 The right to remember does not demand one uncontested history. It protects the evidence and plural institutions required for society to argue about history honestly. proposal
  2. CL-20-04.02 Stage Risk Safeguard proposal
  3. CL-20-04.03 Creation Witnessing, journalism, or documentation is criminalized. Protect observers, journalists, researchers, and secure evidence capture. proposal
  4. CL-20-04.04 Custody Records are destroyed, hidden, or held by one vulnerable institution. Redundant, distributed, authenticated preservation. proposal
  5. CL-20-04.05 Discovery Indexes, search, or AI systems silently suppress the material. Transparent catalogues, interoperable metadata, plural retrieval systems. proposal
  6. CL-20-04.06 Interpretation One official narrative is treated as the only permissible meaning. Access to primary records, competing scholarship, and visible uncertainty. proposal
  7. CL-20-04.07 Correction Fabricated or mistaken records persist without remedy. Versioned corrections that preserve prior states and reasons. proposal
  8. CL-20-04.08 Protection across the historical record lifecycle Stage Risk Safeguard Creation Witnessing, journalism, or documentation is criminalized. Protect observers, journalists, researchers, and secure evidence capture. Custody Records are destroyed, hidden, or held by one vulnerable institution. Redundant, distributed, authenticated preservation. Discovery Indexes, search, or AI systems silently suppress the material. Transparent catalogues, interoperable metadata, plural retrieval systems. Interpretation One official narrative is treated as the only permissible meaning. Access to primary records, competing scholarship, and visible uncertainty. Correction Fabricated or mistaken records persist without remedy. Versioned corrections that preserve prior states and reasons. proposal
4 evidence units
Report-derived argument

Genuine harm can produce overbroad machinery

CL-21-01
As of
Next review
Claim kind
global synthesis

Section summary: Democratic societies face real problems: exploitation, threats, fraud, coordinated manipulation, harassment, and dangerous technical assistance. The drift begins when urgency collapses the distinction between a harmful act and the infrastructure used to discuss, investigate, or detect it. A system designed to find a narrow category of illegal material may

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
  • SRC-SANTA-CLARA-PRINCIPLES Santa Clara Principles on Transparency and Accountability in Content Moderation · Professional or project methodology
Inspect 2 inventoried assertions
  1. CL-21-01.01 Democratic societies face real problems: exploitation, threats, fraud, coordinated manipulation, harassment, and dangerous technical assistance. The drift begins when urgency collapses the distinction between a harmful act and the infrastructure used to discuss, investigate, or detect it. A system designed to find a narrow category of illegal material may require scanning all material. A rule aimed at malicious accounts may make anonymous participation impossible. A duty to remove unlawful speech may encourage deletion of anything controversial. inference or risk
  2. CL-21-01.02 The issue is not whether a protective goal is sincere. It is whether the mechanism is bounded. Systems that observe everyone, place the cost of error on speakers, or make private companies responsible for guessing unsettled law tend to overreach even when no official explicitly orders them to suppress lawful dissent. descriptive assertion
Report-derived argument

Drift follows predictable institutional incentives

CL-21-02
As of
Next review
Claim kind
global synthesis

Section summary: Vague categories: “harmful,” “false,” “extremist,” or “unsafe” lack operational limits. Ruinous penalties: intermediaries remove first because careful review carries asymmetric risk. Short deadlines: context, translation, satire, journalism, and appeal become impossible. Informal pressure: officials achieve indirectly what formal law would struggle to

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-21-02.01 These forces operate across political cultures. The result can be a censorship function without a single institution publicly accepting responsibility for censorship. descriptive assertion
  2. CL-21-02.02 Vague categories: “harmful,” “false,” “extremist,” or “unsafe” lack operational limits. descriptive assertion
  3. CL-21-02.03 Ruinous penalties: intermediaries remove first because careful review carries asymmetric risk. inference or risk
  4. CL-21-02.04 Short deadlines: context, translation, satire, journalism, and appeal become impossible. descriptive assertion
  5. CL-21-02.05 Informal pressure: officials achieve indirectly what formal law would struggle to authorize. descriptive assertion
  6. CL-21-02.06 Identity linkage: enforcement data becomes a general map of political and intellectual activity. descriptive assertion
  7. CL-21-02.07 Emergency renewal: temporary authorities acquire staff, vendors, databases, and new missions. descriptive assertion
Report-derived argument

Institutional friction separates protection from control

CL-21-03
As of
Next review
Claim kind
global synthesis

Section summary: A restrictive proposal does not by itself establish a censorship state. The crucial evidence is what happens around it. Can an independent court narrow the rule? Can journalists inspect implementation? Can affected users see an order and appeal? Can opposition parties and civil society obtain data? Can a provider choose a less intrusive design? Does the

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-21-03.01 A restrictive proposal does not by itself establish a censorship state. The crucial evidence is what happens around it. Can an independent court narrow the rule? Can journalists inspect implementation? Can affected users see an order and appeal? Can opposition parties and civil society obtain data? Can a provider choose a less intrusive design? Does the measure expire unless its value is demonstrated? descriptive assertion
  2. CL-21-03.02 Friction raises the cost of false positives and forces the restricting authority to explain itself. A system in which blocking, surveillance, or identity demands can be triggered automatically, secretly, and at negligible cost is structurally more dangerous than one requiring individualized evidence and adversarial review. inference or risk
Site policy proposal

An anti-drift design for every jurisdiction

CL-21-04
As of
Next review
Claim kind
policy proposal

Section summary: Anti-drift rule: never give an institution a general capability merely because one sympathetic use case exists. Legislation should define the prohibited conduct, evidence threshold, decision-maker, retention period, affected data, appeal, audit, and sunset in the text. Technical systems should minimize collection, separate identity from content, keep

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-OHCHR-ANONYMITY-ENCRYPTION-2015 UN Special Rapporteur — encryption and anonymity in digital communications · Official oversight record
  • SRC-VDEM-DEMOCRACY-REPORTS V-Dem Institute — Democracy Reports · Civil-society or secondary analysis
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
  • SRC-SANTA-CLARA-PRINCIPLES Santa Clara Principles on Transparency and Accountability in Content Moderation · Professional or project methodology
Inspect 3 inventoried assertions
  1. CL-21-04.01 Legislation should define the prohibited conduct, evidence threshold, decision-maker, retention period, affected data, appeal, audit, and sunset in the text. Technical systems should minimize collection, separate identity from content, keep blocklists reviewable, log official interventions, and prevent secondary use. Platforms should disclose government requests and publish enforcement error data. Courts should be able to inspect both policy and system behavior. proposal
  2. CL-21-04.02 Worldwide prevention is strongest before infrastructure is normalized. Once a control plane, identity database, scanning layer, or shared blacklist is embedded across services, later legal reform may limit its use but leave the capability ready for the next emergency. proposal
  3. CL-21-04.03 Anti-drift rule: never give an institution a general capability merely because one sympathetic use case exists. proposal
4 evidence units
Report-derived argument

Censorship is a contest over reality, not only speech

CL-22-01
As of
Next review
Claim kind
global synthesis

Section summary: Authorities censor for different reasons: to protect rulers, enforce religious orthodoxy, maintain colonial hierarchy, prevent revolt, preserve wartime secrecy, guard morality, or respond to real harm. The methods range from licensing and book burning to control of schools, archives, telecommunications, search ranking, and model training. What connects them

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-OHCHR-ICCPR International Covenant on Civil and Political Rights — Article 18 · Primary law or court record
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 2 inventoried assertions
  1. CL-22-01.01 Authorities censor for different reasons: to protect rulers, enforce religious orthodoxy, maintain colonial hierarchy, prevent revolt, preserve wartime secrecy, guard morality, or respond to real harm. The methods range from licensing and book burning to control of schools, archives, telecommunications, search ranking, and model training. What connects them is the concentration of authority to decide which knowledge can circulate. descriptive assertion
  2. CL-22-01.02 This wider view matters because a society can leave formal speech rights intact while manipulating the inputs to thought. If books are unavailable, archives closed, search results distorted, or AI systems trained on a sanitized record, citizens may be free to speak but deprived of the evidence needed to form and test their views. inference or risk
Report-derived argument

Suppression survives by changing form

CL-22-02
As of
Next review
Claim kind
global synthesis

Section summary: When copying is scarce, authorities control presses. When radio crosses borders, they jam signals. When the web distributes publication, they target hosting, payment, search, identity, and routing. When encryption hides content, they seek endpoints, metadata, account registration, or protocol fingerprints. When users turn to AI, control can move into

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
  • SRC-UNESCO-MEMORY-WORLD UNESCO — Memory of the World Programme · Professional or project methodology
Inspect 2 inventoried assertions
  1. CL-22-02.01 When copying is scarce, authorities control presses. When radio crosses borders, they jam signals. When the web distributes publication, they target hosting, payment, search, identity, and routing. When encryption hides content, they seek endpoints, metadata, account registration, or protocol fingerprints. When users turn to AI, control can move into datasets, retrieval indexes, alignment policies, and generated narratives. descriptive assertion
  2. CL-22-02.02 The historical lesson is not technological determinism. New tools can decentralize knowledge and also create new chokepoints. The direction depends on ownership, interoperability, law, incentives, and whether users retain practical alternatives. descriptive assertion
Report-derived argument

Why censorship fails and still causes lasting damage

CL-22-03
As of
Next review
Claim kind
global synthesis

Section summary: Banned ideas frequently survive through hidden copies, oral memory, foreign presses, diaspora networks, private libraries, encoded language, and later archival recovery. That persistence can create a comforting myth that censorship is ultimately futile. The human cost shows otherwise. Suppression can imprison writers, destroy careers, erase witnesses,

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-22-03.01 Banned ideas frequently survive through hidden copies, oral memory, foreign presses, diaspora networks, private libraries, encoded language, and later archival recovery. That persistence can create a comforting myth that censorship is ultimately futile. The human cost shows otherwise. Suppression can imprison writers, destroy careers, erase witnesses, fragment trust, prevent scientific work, and make generations inherit a distorted record before the truth re-emerges. descriptive assertion
  2. CL-22-03.02 Success should therefore not be measured only by whether an idea eventually survives. Cognitive liberty concerns the people denied access in their own time, the questions never asked, the evidence lost, and the social knowledge that could not form. descriptive assertion
Site policy proposal

The durable lessons for a networked world

CL-22-04
As of
Next review
Claim kind
policy proposal

Section summary: Keep publication and access possible through multiple independent channels. Protect anonymity, encryption, and private reading from generalized suspicion. Preserve records in different institutions, formats, and jurisdictions. Make restrictions visible, specific, appealable, and temporary. Protect libraries, universities, journalists, archivists, and

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-UNESCO-MEMORY-WORLD UNESCO — Memory of the World Programme · Professional or project methodology
  • SRC-VDEM-DEMOCRACY-REPORTS V-Dem Institute — Democracy Reports · Civil-society or secondary analysis
  • SRC-OHCHR-ICCPR International Covenant on Civil and Political Rights — Article 18 · Primary law or court record
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 8 inventoried assertions
  1. CL-22-04.01 The worldwide goal is not an information environment with no rules. It is one in which no authority can convert a temporary claim of necessity into permanent ownership of human memory and inquiry. proposal
  2. CL-22-04.02 Keep publication and access possible through multiple independent channels. proposal
  3. CL-22-04.03 Protect anonymity, encryption, and private reading from generalized suspicion. proposal
  4. CL-22-04.04 Preserve records in different institutions, formats, and jurisdictions. proposal
  5. CL-22-04.05 Make restrictions visible, specific, appealable, and temporary. proposal
  6. CL-22-04.06 Protect libraries, universities, journalists, archivists, and translators as civic infrastructure. proposal
  7. CL-22-04.07 Teach source comparison and historical uncertainty rather than replacing one orthodoxy with another. proposal
  8. CL-22-04.08 Resist global technical standards that make censorship cheap to deploy everywhere. proposal
4 evidence units
Report-derived argument

Information can leave territory without escaping power

CL-23-01
As of
Next review
Claim kind
global synthesis

Section summary: When newsrooms close, archives are seized, and local publishing becomes dangerous, journalists and cultural institutions often relocate. Digital distribution lets them continue reaching domestic and international audiences. But exile does not create automatic safety. States can threaten relatives, conduct online harassment, pursue spyware operations,

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-23-01.01 When newsrooms close, archives are seized, and local publishing becomes dangerous, journalists and cultural institutions often relocate. Digital distribution lets them continue reaching domestic and international audiences. But exile does not create automatic safety. States can threaten relatives, conduct online harassment, pursue spyware operations, manipulate international policing systems, pressure host governments, and exploit platform reporting tools. inference or risk
  2. CL-23-01.02 Cognitive liberty therefore has a cross-border dimension. A person’s ability to receive independent information inside one country may depend on the legal and technical safety of a journalist, server, archive, or translator located elsewhere. inference or risk
Report-derived argument

Diaspora institutions preserve more than news

CL-23-02
As of
Next review
Claim kind
global synthesis

Section summary: Exile networks maintain current reporting, but they also preserve language, books, oral history, documentary evidence, artistic memory, and the institutional knowledge needed for future accountability. In conflict settings, videos and posts may be removed by automated moderation, lost when accounts disappear, or altered without provenance. Dedicated

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-SYRIAN-ARCHIVE Syrian Archive — preservation and verification programme · Professional or project methodology
  • SRC-UNESCO-MEMORY-WORLD UNESCO — Memory of the World Programme · Professional or project methodology
Inspect 2 inventoried assertions
  1. CL-23-02.01 Exile networks maintain current reporting, but they also preserve language, books, oral history, documentary evidence, artistic memory, and the institutional knowledge needed for future accountability. In conflict settings, videos and posts may be removed by automated moderation, lost when accounts disappear, or altered without provenance. Dedicated archives can verify, catalogue, and retain these materials for journalism, scholarship, families, and courts. inference or risk
  2. CL-23-02.02 The work requires careful boundaries. Public preservation can endanger sources. A platform copy may lack the context necessary for evidentiary use. Responsible systems need access controls, chain-of-custody records, redaction authority, preservation policies, and transparent correction practices. inference or risk
Report-derived argument

The global support problem is structural

CL-23-03
As of
Next review
Claim kind
global synthesis

Section summary: Emergency grants can help a newsroom escape immediate danger but do not create a durable institution. Exile media need stable legal status, banking, payment processing, insurance, hosting, security expertise, audience research, and long-term editorial capacity. They also need protection against indiscriminate sanctions and compliance systems that treat

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-23-03.01 Emergency grants can help a newsroom escape immediate danger but do not create a durable institution. Exile media need stable legal status, banking, payment processing, insurance, hosting, security expertise, audience research, and long-term editorial capacity. They also need protection against indiscriminate sanctions and compliance systems that treat cross-border journalism as an unexplained financial risk. inference or risk
  2. CL-23-03.02 Platforms should establish escalation paths for verified exiled media and human-rights archives so coordinated mass reporting does not automatically remove evidence or silence a newsroom. Host governments should recognize transnational repression, protect family contacts, and prevent deportation or forced return to danger. inference or risk
Site policy proposal

A worldwide preservation compact

CL-23-04
As of
Next review
Claim kind
policy proposal

Section summary: What resilient support requires Layer Commitment Human safety Emergency visas, source protection, family-risk assessment, and response to cross-border threats. Financial continuity Rights-aware compliance, transparent funding, and multi-year core support. Technical resilience Secure communication, redundant hosting, mirrors, preservation, and incident

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-23-04.01 Protecting information in exile is not charity for one nationality. It is part of a worldwide resilience system that prevents territorial censorship from becoming global erasure. proposal
  2. CL-23-04.02 Layer Commitment proposal
  3. CL-23-04.03 Human safety Emergency visas, source protection, family-risk assessment, and response to cross-border threats. proposal
  4. CL-23-04.04 Financial continuity Rights-aware compliance, transparent funding, and multi-year core support. proposal
  5. CL-23-04.05 Technical resilience Secure communication, redundant hosting, mirrors, preservation, and incident response. proposal
  6. CL-23-04.06 Platform process Human review, notice, restoration, and protection from coordinated abuse of reporting systems. proposal
  7. CL-23-04.07 Archival integrity Provenance, versioning, access controls, redaction authority, and geographically distributed custody. proposal
  8. CL-23-04.08 Editorial trust Clear corrections, funding disclosure, separation of reporting from advocacy, and continued connection to domestic audiences. proposal
  9. CL-23-04.09 What resilient support requires Layer Commitment Human safety Emergency visas, source protection, family-risk assessment, and response to cross-border threats. Financial continuity Rights-aware compliance, transparent funding, and multi-year core support. Technical resilience Secure communication, redundant hosting, mirrors, preservation, and incident response. Platform process Human review, notice, restoration, and protection from coordinated abuse of reporting systems. Archival integrity Provenance, versioning, access controls, redaction authority, and geographically distributed custody. Editorial trust Clear corrections, funding disclosure, separation of reporting from advocacy, and continued connection to domestic audiences. proposal
4 evidence units
Report-derived argument

A censorship stack has multiple owners

CL-24-01
As of
Next review
Claim kind
global synthesis

Section summary: One agency may define prohibited content, another operate surveillance, private network companies host filtering equipment, platforms enforce local rules, banks cut off organizations, schools repeat the official narrative, and automated systems identify targets. No participant needs to control the whole system. Integration produces the effect. This

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-24-01.01 One agency may define prohibited content, another operate surveillance, private network companies host filtering equipment, platforms enforce local rules, banks cut off organizations, schools repeat the official narrative, and automated systems identify targets. No participant needs to control the whole system. Integration produces the effect. inference or risk
  2. CL-24-01.02 This distributed ownership can obscure accountability. A publisher is blocked because an agency sent an order; the agency cites law; the network operator cites compliance; the platform cites terms; the bank cites risk. A person loses practical access or voice while every institution claims it merely followed another layer. inference or risk
Report-derived argument

Law turns technical capability into routine power

CL-24-02
As of
Next review
Claim kind
global synthesis

Section summary: Technical systems do not become censorship infrastructure solely because they can classify or route traffic. The critical transition occurs when broad legal categories, weak review, and centralized command make political use ordinary. Foreign-agent labels, extremism rules, wartime speech offenses, data retention, localization, and licensing can create a

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-24-02.01 Technical systems do not become censorship infrastructure solely because they can classify or route traffic. The critical transition occurs when broad legal categories, weak review, and centralized command make political use ordinary. Foreign-agent labels, extremism rules, wartime speech offenses, data retention, localization, and licensing can create a legal environment in which independent publication is costly even before a block is issued. descriptive assertion
  2. CL-24-02.02 Once control hardware is installed inline with ordinary communications, enforcement can shift from requests to direct execution. The state no longer needs to persuade thousands of independent providers individually. That reduction in institutional friction is a major worldwide warning sign. inference or risk
Report-derived argument

AI lowers the cost of suspicion

CL-24-03
As of
Next review
Claim kind
global synthesis

Section summary: Manual censorship has natural limits: people must read, classify, translate, and prioritize. AI can scan text, images, video, social graphs, and trends at a scale that makes broad categories operational. Predictive systems may flag emerging “tension,” map networks, or identify visual symbols. The danger is not perfect omniscience; imperfect classifiers can

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-24-03.01 Manual censorship has natural limits: people must read, classify, translate, and prioritize. AI can scan text, images, video, social graphs, and trends at a scale that makes broad categories operational. Predictive systems may flag emerging “tension,” map networks, or identify visual symbols. The danger is not perfect omniscience; imperfect classifiers can still chill a population when the consequences of a false positive are severe. inference or risk
  2. CL-24-03.02 Automated monitoring also permits low-cost experimentation. Authorities can adjust taxonomies, add targets, and compare behavior without publicly changing a law. A rights-preserving system must therefore audit model objectives, labels, error rates, data sources, retention, and downstream consequences—not only the text of the statute. descriptive assertion
Site policy proposal

How the world can stop the stack from forming

CL-24-04
As of
Next review
Claim kind
policy proposal

Section summary: Prevent one regulator from controlling both policy and technical execution without independent authorization. Keep telecommunications diverse and prohibit secret inline control of general traffic. Require public legal definitions and published blocking or data-demand statistics. Protect independent media from financial punishment unrelated to proven

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-24-04.01 Studying an integrated censorship stack is most useful as prevention. Once every layer supports the others, reform requires changing law, infrastructure, markets, institutions, and public fear at the same time. proposal
  2. CL-24-04.02 Prevent one regulator from controlling both policy and technical execution without independent authorization. proposal
  3. CL-24-04.03 Keep telecommunications diverse and prohibit secret inline control of general traffic. proposal
  4. CL-24-04.04 Require public legal definitions and published blocking or data-demand statistics. proposal
  5. CL-24-04.05 Protect independent media from financial punishment unrelated to proven unlawful conduct. proposal
  6. CL-24-04.06 Audit AI monitoring for purpose, data, error, targeting, and appeal. proposal
  7. CL-24-04.07 Preserve encryption, anonymity, foreign sources, and alternative distribution channels. proposal
  8. CL-24-04.08 Give courts and civil society access to evidence before emergency powers become normal. proposal
4 evidence units
Report-derived argument

Monitor systems, not national stereotypes

CL-25-01
As of
Next review
Claim kind
global synthesis

Section summary: Censorship risk cannot be inferred from a country name, region, ideology, or single controversial law. The same tool can serve a narrow protective purpose in one setting and become a general control mechanism in another. Analysis should therefore ask who can trigger the power, what evidence is required, how cheaply it scales, what data it creates, whether

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-25-01.01 Censorship risk cannot be inferred from a country name, region, ideology, or single controversial law. The same tool can serve a narrow protective purpose in one setting and become a general control mechanism in another. Analysis should therefore ask who can trigger the power, what evidence is required, how cheaply it scales, what data it creates, whether alternatives remain, and whether independent institutions can reverse an error. inference or risk
  2. CL-25-01.02 This approach also prevents complacency. Democracies can build dangerous infrastructure under sincere safety rationales, while some restrictive proposals may be narrowed or defeated by courts and civil society. Trajectory and institutional response matter more than labels. inference or risk
Report-derived argument

Four domains reveal the direction of travel

CL-25-02
As of
Next review
Claim kind
global synthesis

Section summary: Legal expansion Vague offenses, emergency powers, publisher licensing, foreign-agent rules, and criminalized “false information.” Institutional capture Politicized regulators, weakened courts, secret orders, selective prosecution, and financial pressure. Technical control Centralized gateways, deep packet inspection, identity mandates, encryption

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 6 inventoried assertions
  1. CL-25-02.01 Vague offenses, emergency powers, publisher licensing, foreign-agent rules, and criminalized “false information.” descriptive assertion
  2. CL-25-02.02 Politicized regulators, weakened courts, secret orders, selective prosecution, and financial pressure. descriptive assertion
  3. CL-25-02.03 Centralized gateways, deep packet inspection, identity mandates, encryption restrictions, and AI monitoring. descriptive assertion
  4. CL-25-02.04 Ruinous liability, short takedown deadlines, app-store pressure, hosting denial, and informal government demands. descriptive assertion
  5. CL-25-02.05 One indicator may reflect a genuine and bounded response. Multiple domains moving together indicate a system capable of dynamic, predictive, and punitive information control. inference or risk
  6. CL-25-02.06 Legal expansion Vague offenses, emergency powers, publisher licensing, foreign-agent rules, and criminalized “false information.” Institutional capture Politicized regulators, weakened courts, secret orders, selective prosecution, and financial pressure. Technical control Centralized gateways, deep packet inspection, identity mandates, encryption restrictions, and AI monitoring. Intermediary coercion Ruinous liability, short takedown deadlines, app-store pressure, hosting denial, and informal government demands. descriptive assertion
Report-derived argument

The most important measure is friction

CL-25-03
As of
Next review
Claim kind
global synthesis

Section summary: Rights survive when exercising power is costly. A targeted warrant requires evidence. A public order can be challenged. A regulator must explain its reasoning. A platform bears a cost for wrongful removal. An independent network can refuse an improper demand. A legislature must vote again when an authority expires. Censorship becomes systemic when those

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-OONI-MEASUREMENT OONI — open measurement of internet censorship · Professional or project methodology
  • SRC-SANTA-CLARA-PRINCIPLES Santa Clara Principles on Transparency and Accountability in Content Moderation · Professional or project methodology
Inspect 2 inventoried assertions
  1. CL-25-03.01 Rights survive when exercising power is costly. A targeted warrant requires evidence. A public order can be challenged. A regulator must explain its reasoning. A platform bears a cost for wrongful removal. An independent network can refuse an improper demand. A legislature must vote again when an authority expires. descriptive assertion
  2. CL-25-03.02 Censorship becomes systemic when those costs approach zero: a classifier flags millions of items, a blocklist updates invisibly, a national gateway executes centrally, and affected people never learn why. Analysts should therefore track the “cost to trigger” and “cost to correct” each intervention. Low-cost restriction combined with high-cost appeal is a critical warning. inference or risk
Site policy proposal

A practical worldwide monitoring checklist

CL-25-04
As of
Next review
Claim kind
policy proposal

Section summary: Questions to review on a regular cycle Question Protective evidence Escalation evidence Are definitions narrow? Objective elements, intent, harm, and judicial interpretation. Elastic terms applied to peaceful criticism or research. Can institutions resist? Courts, opposition, regulators, and media obtain records and reverse overreach. Orders pass without

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-SANTA-CLARA-PRINCIPLES Santa Clara Principles on Transparency and Accountability in Content Moderation · Professional or project methodology
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
  • SRC-VDEM-DEMOCRACY-REPORTS V-Dem Institute — Democracy Reports · Civil-society or secondary analysis
  • SRC-FREEDOM-HOUSE-FOTN Freedom House — Freedom on the Net · Civil-society or secondary analysis
  • SRC-OONI-MEASUREMENT OONI — open measurement of internet censorship · Professional or project methodology
Inspect 8 inventoried assertions
  1. CL-25-04.01 A public score can be useful only when evidence is current, methods are transparent, uncertainty is shown, and one country is not treated as the template for all others. This release therefore provides the framework, not a stale global ranking. proposal
  2. CL-25-04.02 Question Protective evidence Escalation evidence proposal
  3. CL-25-04.03 Are definitions narrow? Objective elements, intent, harm, and judicial interpretation. Elastic terms applied to peaceful criticism or research. proposal
  4. CL-25-04.04 Can institutions resist? Courts, opposition, regulators, and media obtain records and reverse overreach. Orders pass without evidence, review, or published reasoning. proposal
  5. CL-25-04.05 Is infrastructure centralized? Diverse providers and independent control planes. One authority can throttle, filter, identify, or disconnect at scale. proposal
  6. CL-25-04.06 Are errors visible? Notice, error statistics, restoration, and compensation. Silent demotion, secret lists, and no appeal. proposal
  7. CL-25-04.07 Does the power end? Deletion, sunset, dismantlement, and proof of continuing need. Permanent data, vendors, and renewed emergency mandates. proposal
  8. CL-25-04.08 Questions to review on a regular cycle Question Protective evidence Escalation evidence Are definitions narrow? Objective elements, intent, harm, and judicial interpretation. Elastic terms applied to peaceful criticism or research. Can institutions resist? Courts, opposition, regulators, and media obtain records and reverse overreach. Orders pass without evidence, review, or published reasoning. Is infrastructure centralized? Diverse providers and independent control planes. One authority can throttle, filter, identify, or disconnect at scale. Are errors visible? Notice, error statistics, restoration, and compensation. Silent demotion, secret lists, and no appeal. Does the power end? Deletion, sunset, dismantlement, and proof of continuing need. Permanent data, vendors, and renewed emergency mandates. proposal
4 evidence units
Evidence + editorial analysis

The right protects the process of seeking as well as speaking

CL-26-01
As of
Next review
Claim kind
global synthesis

Section summary: Expression law often focuses on the speaker, but cognitive liberty also protects the listener, reader, researcher, and person still deciding what they believe. A rule that leaves speech formally legal while blocking discovery, criminalizing receipt, or identifying every reader can damage the same underlying freedom. International human-rights principles

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-OHCHR-ICCPR International Covenant on Civil and Political Rights — Article 18 · Primary law or court record
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 2 inventoried assertions
  1. CL-26-01.01 Expression law often focuses on the speaker, but cognitive liberty also protects the listener, reader, researcher, and person still deciding what they believe. A rule that leaves speech formally legal while blocking discovery, criminalizing receipt, or identifying every reader can damage the same underlying freedom. descriptive assertion
  2. CL-26-01.02 International human-rights principles recognize seeking, receiving, and imparting information across borders. Restrictions remain possible, but they must be treated as exceptions. The state should not reverse that relationship by requiring citizens to prove that every controversial question is safe before access is allowed. descriptive assertion
Report-derived argument

The harm category must be operational, not rhetorical

CL-26-02
As of
Next review
Claim kind
global synthesis

Section summary: “Harmful speech” can refer to very different phenomena: a true threat directed at a person, intentional incitement likely to produce violence, defamation, exploitation, fraud, harassment, political propaganda, inaccurate medical claims, offensive art, or criticism of institutions. Applying one automated rule to all of them guarantees error. A

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
  • SRC-OHCHR-RABAT-PLAN OHCHR — Rabat Plan of Action · Professional or project methodology
Inspect 2 inventoried assertions
  1. CL-26-02.01 “Harmful speech” can refer to very different phenomena: a true threat directed at a person, intentional incitement likely to produce violence, defamation, exploitation, fraud, harassment, political propaganda, inaccurate medical claims, offensive art, or criticism of institutions. Applying one automated rule to all of them guarantees error. descriptive assertion
  2. CL-26-02.02 A rights-preserving system defines the protected interest, the prohibited act, the required mental state, the causal connection, and the available remedy. It distinguishes disagreement from threat, analysis from advocacy, and advocacy from operational coordination. The more severe the penalty or prior restraint, the stronger the evidence and review should be. descriptive assertion
Evidence + editorial analysis

Use a six-factor threshold before coercive restriction

CL-26-03
As of
Next review
Claim kind
global synthesis

Section summary: Context What social conditions and vulnerabilities make harm more or less likely? Speaker What authority, influence, or relationship does the speaker hold? Intent Was harm advocated deliberately, or is the material analytical, satirical, or mistaken? Content and form Is there a direct call, threat, or operational instruction? Extent What is the reach,

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-OHCHR-RABAT-PLAN OHCHR — Rabat Plan of Action · Professional or project methodology
  • SRC-SANTA-CLARA-PRINCIPLES Santa Clara Principles on Transparency and Accountability in Content Moderation · Professional or project methodology
Inspect 7 inventoried assertions
  1. CL-26-03.01 These factors do not mechanically decide every case. They prevent a broad label from doing all the legal work. descriptive assertion
  2. CL-26-03.02 Context What social conditions and vulnerabilities make harm more or less likely? descriptive assertion
  3. CL-26-03.03 Speaker What authority, influence, or relationship does the speaker hold? descriptive assertion
  4. CL-26-03.04 Intent Was harm advocated deliberately, or is the material analytical, satirical, or mistaken? descriptive assertion
  5. CL-26-03.05 Content and form Is there a direct call, threat, or operational instruction? descriptive assertion
  6. CL-26-03.06 Extent What is the reach, repetition, targeting, and distribution method? descriptive assertion
  7. CL-26-03.07 Likelihood How probable and proximate is the alleged harm? descriptive assertion
Site policy proposal

A worldwide due-process floor

CL-26-04
As of
Next review
Claim kind
policy proposal

Section summary: Publish the rule and define its terms before enforcement. Require a reasoned decision tied to specific content and law. Give users and publishers notice unless a narrowly justified delay is necessary. Provide fast human appeal, restoration, and correction of records. Report government demands and error rates in comparable categories. Protect encryption,

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-OHCHR-RABAT-PLAN OHCHR — Rabat Plan of Action · Professional or project methodology
  • SRC-SANTA-CLARA-PRINCIPLES Santa Clara Principles on Transparency and Accountability in Content Moderation · Professional or project methodology
  • SRC-OHCHR-ICCPR International Covenant on Civil and Political Rights — Article 18 · Primary law or court record
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 8 inventoried assertions
  1. CL-26-04.01 The objective is not legal uniformity across the world. It is a minimum protection against systems that use the language of safety to make political or intellectual conformity the default. proposal
  2. CL-26-04.02 Publish the rule and define its terms before enforcement. proposal
  3. CL-26-04.03 Require a reasoned decision tied to specific content and law. proposal
  4. CL-26-04.04 Give users and publishers notice unless a narrowly justified delay is necessary. proposal
  5. CL-26-04.05 Provide fast human appeal, restoration, and correction of records. proposal
  6. CL-26-04.06 Report government demands and error rates in comparable categories. proposal
  7. CL-26-04.07 Protect encryption, anonymity, journalism, research, satire, and historical documentation from collateral enforcement. proposal
  8. CL-26-04.08 Prefer counterspeech, labels, user controls, and targeted remedies when they can address the harm. proposal
4 evidence units
Evidence + editorial analysis

AI moves mediation from ranking to synthesis

CL-27-01
As of
Next review
Claim kind
global synthesis

Section summary: Traditional search already ranks and filters, but the user can often inspect several sources and notice disagreement. An AI assistant selects passages, weighs them, generates a narrative, and presents the result in a conversational voice. That convenience changes the locus of judgment. The assistant can become the place where a person frames the question,

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-27-01.01 Traditional search already ranks and filters, but the user can often inspect several sources and notice disagreement. An AI assistant selects passages, weighs them, generates a narrative, and presents the result in a conversational voice. That convenience changes the locus of judgment. The assistant can become the place where a person frames the question, explores uncertainty, tests values, drafts speech, and decides what evidence exists. inference or risk
  2. CL-27-01.02 This makes prompt logs and long-running conversations unusually sensitive. They may reveal tentative beliefs, private problems, emotional states, professional research, and questions the user would never publish. Treating those records as ordinary product telemetry misunderstands their cognitive role. inference or risk
Evidence + editorial analysis

Soft censorship and sycophancy distort in opposite directions

CL-27-02
As of
Next review
Claim kind
global synthesis

Section summary: Hard censorship tells the user that an answer was withheld. Soft censorship produces an answer while hiding relevant evidence, framing the issue as settled, or substituting a safer but different question. Sycophancy creates another distortion: the model mirrors the user’s assumptions, ideology, or desired conclusion instead of providing independent

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-27-02.01 Hard censorship tells the user that an answer was withheld. Soft censorship produces an answer while hiding relevant evidence, framing the issue as settled, or substituting a safer but different question. Sycophancy creates another distortion: the model mirrors the user’s assumptions, ideology, or desired conclusion instead of providing independent friction. descriptive assertion
  2. CL-27-02.02 One mechanism narrows the answer to satisfy policy; the other bends the answer to satisfy the user. Both can weaken autonomy when the adaptation is invisible. A liberty-preserving assistant should explain uncertainty, surface competing credible accounts, distinguish evidence from interpretation, and make personalization controllable rather than covert. descriptive assertion
Evidence + editorial analysis

Geographic and linguistic fragmentation can localize reality

CL-27-03
As of
Next review
Claim kind
global synthesis

Section summary: AI services operate under different laws, training corpora, market incentives, and cultural assumptions. The same question may produce different refusals, omissions, or moral framing depending on language, region, account state, or provider. Some variation is necessary—law and context differ—but invisible political localization can make users believe a

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-27-03.01 AI services operate under different laws, training corpora, market incentives, and cultural assumptions. The same question may produce different refusals, omissions, or moral framing depending on language, region, account state, or provider. Some variation is necessary—law and context differ—but invisible political localization can make users believe a generated answer is a neutral summary of the world. inference or risk
  2. CL-27-03.02 Cross-lingual audits, source comparisons, and public policy-change records are therefore essential. A global system should not provide weaker epistemic transparency merely because a language has fewer evaluation resources or a population has fewer alternative providers. descriptive assertion
Site policy proposal

A bill of rights for AI-mediated thought

CL-27-04
As of
Next review
Claim kind
policy proposal

Section summary: User rights for AI assistants Right Required system behavior Private inquiry Short retention, no unrelated profiling, and a practical local or anonymous mode. Source access Identifiable evidence, retrieval context, and the ability to inspect original material. Visible boundaries Explain whether law, product policy, uncertainty, or technical limits shaped

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 10 inventoried assertions
  1. CL-27-04.01 The assistant should expand the user’s capacity to reason, not quietly become the unchallengeable author of the user’s information environment. proposal
  2. CL-27-04.02 Right Required system behavior proposal
  3. CL-27-04.03 Private inquiry Short retention, no unrelated profiling, and a practical local or anonymous mode. proposal
  4. CL-27-04.04 Source access Identifiable evidence, retrieval context, and the ability to inspect original material. proposal
  5. CL-27-04.05 Visible boundaries Explain whether law, product policy, uncertainty, or technical limits shaped the answer. proposal
  6. CL-27-04.06 Intent distinction Answer lawful historical, analytical, defensive, and critical inquiry while restricting narrowly harmful capability uplift. proposal
  7. CL-27-04.07 Memory control View, correct, delete, export, and disable persistent context. proposal
  8. CL-27-04.08 Plurality Interoperability and practical access to models with different governance choices. proposal
  9. CL-27-04.09 Contestability Appeal consequential refusals, suspensions, or profiles and correct false inferences. proposal
  10. CL-27-04.10 User rights for AI assistants Right Required system behavior Private inquiry Short retention, no unrelated profiling, and a practical local or anonymous mode. Source access Identifiable evidence, retrieval context, and the ability to inspect original material. Visible boundaries Explain whether law, product policy, uncertainty, or technical limits shaped the answer. Intent distinction Answer lawful historical, analytical, defensive, and critical inquiry while restricting narrowly harmful capability uplift. Memory control View, correct, delete, export, and disable persistent context. Plurality Interoperability and practical access to models with different governance choices. Contestability Appeal consequential refusals, suspensions, or profiles and correct false inferences. proposal
4 evidence units
Evidence + editorial analysis

Surveillance is most efficient when it becomes anticipated

CL-28-01
As of
Next review
Claim kind
global synthesis

Section summary: Continuous observation is expensive. Anticipated observation is scalable. When rules are vague and consequences severe, people plan for the worst case: they avoid a search term, delete a draft, use safer language, stop contacting a source, or remain silent in a group. The censor’s work is transferred into ordinary decision-making. This is why privacy cannot

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-28-01.01 Continuous observation is expensive. Anticipated observation is scalable. When rules are vague and consequences severe, people plan for the worst case: they avoid a search term, delete a draft, use safer language, stop contacting a source, or remain silent in a group. The censor’s work is transferred into ordinary decision-making. descriptive assertion
  2. CL-28-01.02 This is why privacy cannot be reduced to whether a human official actually opened a specific record. The existence of persistent logs, automated risk scoring, or unclear disclosure rules can alter behavior before any enforcement occurs. inference or risk
Evidence + editorial analysis

Silence manufactures false consensus

CL-28-02
As of
Next review
Claim kind
global synthesis

Section summary: Preference falsification occurs when people publicly perform an acceptable belief while privately holding another. The spiral of silence compounds the effect: people observe the absence of dissent, infer that they are isolated, and remain quiet. Public agreement can therefore become an artifact of fear rather than a measurement of conviction. This damages

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-28-02.01 Preference falsification occurs when people publicly perform an acceptable belief while privately holding another. The spiral of silence compounds the effect: people observe the absence of dissent, infer that they are isolated, and remain quiet. Public agreement can therefore become an artifact of fear rather than a measurement of conviction. descriptive assertion
  2. CL-28-02.02 This damages collective knowledge. Institutions receive distorted feedback, errors persist, and sudden political shifts appear inexplicable because private disagreement was never visible. AI systems trained on public behavior may then learn the falsified consensus and reproduce it as a description of what “people believe.” inference or risk
Report-derived argument

Friction and flooding reshape attention without an explicit ban

CL-28-03
As of
Next review
Claim kind
global synthesis

Section summary: A censored site may load slowly, require repeated workarounds, disappear from recommendations, or become harder to search. Each obstacle is small, but over time people move toward easier sources. Flooding achieves a similar result by overwhelming limited attention with distraction, repetition, or manufactured consensus. These mechanisms are difficult to

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-28-03.01 A censored site may load slowly, require repeated workarounds, disappear from recommendations, or become harder to search. Each obstacle is small, but over time people move toward easier sources. Flooding achieves a similar result by overwhelming limited attention with distraction, repetition, or manufactured consensus. inference or risk
  2. CL-28-03.02 These mechanisms are difficult to contest because no single item is necessarily prohibited. The intervention works at the level of attention and convenience. Audits should therefore examine reach, latency, recommendation distribution, search visibility, and coordinated amplification—not only formal takedown counts. descriptive assertion
Site policy proposal

Psychological resilience needs privacy and common knowledge

CL-28-04
As of
Next review
Claim kind
policy proposal

Section summary: Guarantee confidential spaces for reading, research, counseling, journalism, and association. Use narrow rules so people can know what is actually prohibited. Publish enforcement data and successful appeals to reduce uncertainty. Protect independent media and forums that reveal dissent is not solitary. Teach source evaluation, propaganda recognition, and

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-28-04.01 The long-term aim is not merely helping people cope with censorship. It is preventing institutions from making fear and self-policing normal features of intellectual life. proposal
  2. CL-28-04.02 Guarantee confidential spaces for reading, research, counseling, journalism, and association. proposal
  3. CL-28-04.03 Use narrow rules so people can know what is actually prohibited. proposal
  4. CL-28-04.04 Publish enforcement data and successful appeals to reduce uncertainty. proposal
  5. CL-28-04.05 Protect independent media and forums that reveal dissent is not solitary. proposal
  6. CL-28-04.06 Teach source evaluation, propaganda recognition, and the difference between popularity and truth. proposal
  7. CL-28-04.07 Support local cultural forms—literature, satire, art, private discussion—that sustain identity under pressure. proposal
  8. CL-28-04.08 Limit behavioral inference and prevent inquiry profiles from affecting unrelated opportunities. proposal
  9. CL-28-04.09 Resilience also depends on common knowledge: people need reliable ways to learn that others share concerns, that lawful dissent exists, and that institutional boundaries can be challenged. Privacy protects the individual while independent public forums prevent isolation from becoming manufactured consent. proposal
4 evidence units
Report-derived argument

A national censorship system depends on many commercial layers

CL-29-01
As of
Next review
Claim kind
global synthesis

Section summary: States define the rules, but implementation can involve routers, filtering databases, traffic classifiers, interception hardware, biometric identity, local telecommunications operators, cloud services, app stores, spyware, and analytical software. These components may originate in several countries and pass through distributors or integrators before

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-29-01.01 States define the rules, but implementation can involve routers, filtering databases, traffic classifiers, interception hardware, biometric identity, local telecommunications operators, cloud services, app stores, spyware, and analytical software. These components may originate in several countries and pass through distributors or integrators before deployment. inference or risk
  2. CL-29-01.02 The resulting accountability gap is predictable. A manufacturer says it sold a general network product. A reseller says it followed local law. An operator says it was required to install the system. The state treats vendor details as security information. Without procurement records and independent technical evidence, affected users cannot identify the decision chain. descriptive assertion
Report-derived argument

Dual-use capability is real and not a complete defense

CL-29-02
As of
Next review
Claim kind
global synthesis

Section summary: Deep packet inspection can detect malware, manage congestion, enforce enterprise policy, or protect a service from abuse. The same classification and intervention capability can identify a news site, throttle a platform, block an encrypted protocol, or inject a redirect. URL categorization can support parental controls and also suppress health, minority,

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-29-02.01 Deep packet inspection can detect malware, manage congestion, enforce enterprise policy, or protect a service from abuse. The same classification and intervention capability can identify a news site, throttle a platform, block an encrypted protocol, or inject a redirect. URL categorization can support parental controls and also suppress health, minority, religious, or political information. descriptive assertion
  2. CL-29-02.02 Because the legitimate use is real, blanket prohibition may be technically and economically unworkable. But “dual use” cannot end the analysis. Vendors should assess the customer, legal environment, requested categories, network position, remote-control capability, maintenance relationship, and evidence of prior misuse. inference or risk
Report-derived argument

Censorship and surveillance form one feedback loop

CL-29-03
As of
Next review
Claim kind
global synthesis

Section summary: A blocking system sits where it can observe attempted access. Logs may reveal which people sought prohibited information, which tools they used, and which communities share links. Identity systems, mobile registration, device fingerprints, and data-broker records can turn network events into named profiles. Enforcement then teaches users that attempted

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-29-03.01 A blocking system sits where it can observe attempted access. Logs may reveal which people sought prohibited information, which tools they used, and which communities share links. Identity systems, mobile registration, device fingerprints, and data-broker records can turn network events into named profiles. Enforcement then teaches users that attempted inquiry itself carries risk. inference or risk
  2. CL-29-03.02 AI makes this loop more scalable by classifying traffic, images, language, and social patterns. The important audit questions are what labels exist, how they change, where logs go, who can query them, how long they persist, and whether a person can challenge an attribution. descriptive assertion
Site policy proposal

A global vendor accountability framework

CL-29-04
As of
Next review
Claim kind
policy proposal

Section summary: Controls across the technology lifecycle Stage Required safeguard Design Minimize logging, separate functions, require authenticated policy changes, and make abuse detectable. Sales Conduct rights due diligence on customer, jurisdiction, end use, and requested capabilities. Contract Prohibit political filtering, require audit rights, disclose resellers, and

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-29-04.01 The worldwide objective is to make repression harder to purchase as a turnkey service and harder for every participant in the supply chain to deny. proposal
  2. CL-29-04.02 Stage Required safeguard proposal
  3. CL-29-04.03 Design Minimize logging, separate functions, require authenticated policy changes, and make abuse detectable. proposal
  4. CL-29-04.04 Sales Conduct rights due diligence on customer, jurisdiction, end use, and requested capabilities. proposal
  5. CL-29-04.05 Contract Prohibit political filtering, require audit rights, disclose resellers, and permit termination for misuse. proposal
  6. CL-29-04.06 Deployment Publish procurement, categories, legal authority, scope, and independent impact assessment where possible. proposal
  7. CL-29-04.07 Operation Maintain tamper-evident logs, public statistics, external measurement, and rapid correction of overblocking. proposal
  8. CL-29-04.08 Remedy Provide notice, appeal, restoration, compensation, and evidence preservation for affected users. proposal
  9. CL-29-04.09 Controls across the technology lifecycle Stage Required safeguard Design Minimize logging, separate functions, require authenticated policy changes, and make abuse detectable. Sales Conduct rights due diligence on customer, jurisdiction, end use, and requested capabilities. Contract Prohibit political filtering, require audit rights, disclose resellers, and permit termination for misuse. Deployment Publish procurement, categories, legal authority, scope, and independent impact assessment where possible. Operation Maintain tamper-evident logs, public statistics, external measurement, and rapid correction of overblocking. Remedy Provide notice, appeal, restoration, compensation, and evidence preservation for affected users. proposal
4 evidence units
Report-derived argument

Resilience and sovereign control answer different questions

CL-30-01
As of
Next review
Claim kind
global synthesis

Section summary: Network resilience asks whether essential communications continue during failure, attack, disaster, or disconnection. A sovereign-control project asks who can direct traffic, define permissible destinations, and compel local infrastructure. The two can overlap, but they should not be treated as identical. Redundant links, secure routing, resilient DNS,

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-30-01.01 Network resilience asks whether essential communications continue during failure, attack, disaster, or disconnection. A sovereign-control project asks who can direct traffic, define permissible destinations, and compel local infrastructure. The two can overlap, but they should not be treated as identical. descriptive assertion
  2. CL-30-01.02 Redundant links, secure routing, resilient DNS, local caching, incident response, and diverse providers can improve continuity. A centralized inline control plane adds the ability to filter and throttle all traffic. If one institution holds both emergency routing authority and political content authority, a cybersecurity tool can become an information-control system with little additional engineering. inference or risk
Report-derived argument

Centralization changes the cost of censorship

CL-30-02
As of
Next review
Claim kind
global synthesis

Section summary: A decentralized network forces a censor to coordinate many providers, tolerate inconsistent implementation, and accept collateral damage. Centralized command lets one authority push a rule to thousands of points, target a protocol, isolate a region, or alter national name resolution. This does not guarantee perfect control, but it reduces the institutional

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-30-02.01 A decentralized network forces a censor to coordinate many providers, tolerate inconsistent implementation, and accept collateral damage. Centralized command lets one authority push a rule to thousands of points, target a protocol, isolate a region, or alter national name resolution. This does not guarantee perfect control, but it reduces the institutional and technical friction that protects users. descriptive assertion
  2. CL-30-02.02 Mandatory data localization and local corporate presence can reinforce the architecture by placing data, equipment, assets, and employees within reach of domestic orders. The issue is not local infrastructure itself; it is whether localization becomes leverage for surveillance, content control, or retaliation without independent process. descriptive assertion
Report-derived argument

Fragmentation harms more than access to news

CL-30-03
As of
Next review
Claim kind
global synthesis

Section summary: Nationally segmented networks can reduce scientific exchange, software security, commerce, education, diaspora contact, cultural continuity, and the ability to compare public claims. Protocol blocking also creates security costs: disabling privacy innovations can expose all users, while pushing people toward unmaintained tools can increase fraud and malware

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-30-03.01 Nationally segmented networks can reduce scientific exchange, software security, commerce, education, diaspora contact, cultural continuity, and the ability to compare public claims. Protocol blocking also creates security costs: disabling privacy innovations can expose all users, while pushing people toward unmaintained tools can increase fraud and malware risk. inference or risk
  2. CL-30-03.02 A fragmented world may produce regionally distinct AI systems, archives, and histories that no longer share sources or standards. Cognitive liberty depends on the ability to cross those boundaries and inspect competing accounts. inference or risk
Site policy proposal

Build security without a censorship switch

CL-30-04
As of
Next review
Claim kind
policy proposal

Section summary: Separate emergency continuity authority from content regulation. Distribute control across independent operators with auditable, time-limited emergency procedures. Use open routing-security and integrity standards rather than secret filtering logic. Preserve encrypted protocols and require evidence before restricting a technical class. Publish network

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-30-04.01 A secure network should remain interoperable and difficult for any single political actor to convert into a national information boundary. proposal
  2. CL-30-04.02 Separate emergency continuity authority from content regulation. proposal
  3. CL-30-04.03 Distribute control across independent operators with auditable, time-limited emergency procedures. proposal
  4. CL-30-04.04 Use open routing-security and integrity standards rather than secret filtering logic. proposal
  5. CL-30-04.05 Preserve encrypted protocols and require evidence before restricting a technical class. proposal
  6. CL-30-04.06 Publish network intervention orders, scope, duration, error, and appeal data. proposal
  7. CL-30-04.07 Protect cross-border research, journalism, cloud access, and diaspora communication. proposal
  8. CL-30-04.08 Test whether resilience goals can be met without observing content or identity. proposal
4 evidence units
Report-derived argument

Resistance changes the medium and the social topology

CL-31-01
As of
Next review
Claim kind
global synthesis

Section summary: Hand-copied manuscripts, clandestine presses, typewritten samizdat, cassettes, foreign radio, photocopies, fax, removable media, websites, encrypted messaging, mirrors, and distributed archives all solve versions of the same problem: a powerful institution controls the normal path from creator to audience. The most resilient response is rarely a stronger

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-OONI-MEASUREMENT OONI — open measurement of internet censorship · Professional or project methodology
  • SRC-TOR-CENSORSHIP Tor Project — censorship circumvention support · Professional or project methodology
Inspect 2 inventoried assertions
  1. CL-31-01.01 Hand-copied manuscripts, clandestine presses, typewritten samizdat, cassettes, foreign radio, photocopies, fax, removable media, websites, encrypted messaging, mirrors, and distributed archives all solve versions of the same problem: a powerful institution controls the normal path from creator to audience. descriptive assertion
  2. CL-31-01.02 The most resilient response is rarely a stronger central broadcaster. It is a network with many copy points, routes, custodians, and communities capable of verifying what they receive. Technological progress increases speed and reach, but it also creates new metadata and chokepoints. descriptive assertion
Report-derived argument

No channel is safe or sufficient by itself

CL-31-02
As of
Next review
Claim kind
global synthesis

Section summary: A physical copy can be seized. A trusted courier can be compromised. A radio signal can be located or jammed. A website can be blocked. A VPN protocol can be fingerprinted. A platform can remove an account. A satellite terminal can become a visible target. A local AI model can contain censored or poisoned training data. Responsible resilience planning uses

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-31-02.01 A physical copy can be seized. A trusted courier can be compromised. A radio signal can be located or jammed. A website can be blocked. A VPN protocol can be fingerprinted. A platform can remove an account. A satellite terminal can become a visible target. A local AI model can contain censored or poisoned training data. inference or risk
  2. CL-31-02.02 Responsible resilience planning uses defense in depth and threat-model specificity. It considers who the adversary is, what metadata is visible, what punishment exists, who can operate the tool, and what happens when a provider disappears. Advice that ignores local consequences can endanger the people it intends to help. inference or risk
Report-derived argument

The recurring principles of survivability

CL-31-03
As of
Next review
Claim kind
global synthesis

Section summary: REDUNDANCY No single deletion destroys the record. DIVERSITY Different channels fail under different attacks. USABILITY A tool protects only people who can use it safely. TRUST Communities need ways to authenticate information and one another. Other durable principles include low bandwidth, offline capability, open formats, local language, distributed

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 5 inventoried assertions
  1. CL-31-03.01 Other durable principles include low bandwidth, offline capability, open formats, local language, distributed governance, documented provenance, and graceful failure. The best system preserves access without forcing every user to become a security specialist. descriptive assertion
  2. CL-31-03.02 REDUNDANCY No single deletion destroys the record. descriptive assertion
  3. CL-31-03.03 DIVERSITY Different channels fail under different attacks. descriptive assertion
  4. CL-31-03.04 USABILITY A tool protects only people who can use it safely. descriptive assertion
  5. CL-31-03.05 TRUST Communities need ways to authenticate information and one another. descriptive assertion
Site policy proposal

Build resilience before it becomes resistance

CL-31-04
As of
Next review
Claim kind
policy proposal

Section summary: Fund independent media, libraries, archives, and translation in ordinary times. Preserve important records in open formats and multiple legal jurisdictions. Protect encryption, interoperability, open-source software, and security research. Teach digital safety and source verification without normalizing constant fear. Maintain offline continuity plans for

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-UNESCO-MEMORY-WORLD UNESCO — Memory of the World Programme · Professional or project methodology
  • SRC-SYRIAN-ARCHIVE Syrian Archive — preservation and verification programme · Professional or project methodology
  • SRC-OONI-MEASUREMENT OONI — open measurement of internet censorship · Professional or project methodology
  • SRC-TOR-CENSORSHIP Tor Project — censorship circumvention support · Professional or project methodology
  • SRC-INTERNET-SOCIETY-FRAGMENTATION Internet Society — internet fragmentation resources · Professional or project methodology
Inspect 9 inventoried assertions
  1. CL-31-04.01 Resistance is most humane when ordinary institutions are already resilient enough that people do not have to choose between ignorance and extreme personal risk. proposal
  2. CL-31-04.02 Fund independent media, libraries, archives, and translation in ordinary times. proposal
  3. CL-31-04.03 Preserve important records in open formats and multiple legal jurisdictions. proposal
  4. CL-31-04.04 Protect encryption, interoperability, open-source software, and security research. proposal
  5. CL-31-04.05 Teach digital safety and source verification without normalizing constant fear. proposal
  6. CL-31-04.06 Maintain offline continuity plans for schools, newsrooms, communities, and public-interest organizations. proposal
  7. CL-31-04.07 Create legal defense and emergency support for journalists, archivists, and technologists. proposal
  8. CL-31-04.08 Measure censorship openly so denial and accidental outages can be distinguished. proposal
  9. CL-31-04.09 Preparedness should be ordinary civic infrastructure, not a last-minute underground project. Redundant hosting, mirrored archives, interoperable tools, multilingual copies, legal support, and low-bandwidth distribution reduce the chance that one order, vendor, border, or network failure can erase access for an entire community. proposal
4 evidence units
Report-derived argument

AI regulation acts on the infrastructure of knowledge

CL-32-01
As of
Next review
Claim kind
global synthesis

Section summary: AI governance can regulate a deployed decision system, a model provider, a training run, a dataset, compute access, an app store, an identity credential, generated content, or the user’s request. Each layer creates different civil-liberties consequences. A rule aimed at discriminatory hiring is not equivalent to a rule requiring every private prompt to be

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-UNESCO-AI-ETHICS UNESCO Recommendation on the Ethics of Artificial Intelligence · Primary law or court record
  • SRC-C2PA-SPEC-2-4 C2PA Specifications 2.4 · Official technical standard
Inspect 2 inventoried assertions
  1. CL-32-01.01 AI governance can regulate a deployed decision system, a model provider, a training run, a dataset, compute access, an app store, an identity credential, generated content, or the user’s request. Each layer creates different civil-liberties consequences. A rule aimed at discriminatory hiring is not equivalent to a rule requiring every private prompt to be logged. A provenance label is not equivalent to refusing all unsigned media. descriptive assertion
  2. CL-32-01.02 Policy should begin by naming the intervention point and explaining why that point is necessary. Otherwise a narrow concern about one use can justify control over the general-purpose tools through which people research, create, criticize, and communicate. descriptive assertion
Report-derived argument

Regulatory convergence can create a global control template

CL-32-02
As of
Next review
Claim kind
global synthesis

Section summary: Large providers often implement one global compliance architecture for many legal systems. Requirements for identity, logging, model registration, output filtering, and rapid removal can therefore become technical defaults beyond the jurisdiction that enacted them. Restrictive governments can copy the vocabulary of safety, authenticity, child protection, or

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
Inspect 2 inventoried assertions
  1. CL-32-02.01 Large providers often implement one global compliance architecture for many legal systems. Requirements for identity, logging, model registration, output filtering, and rapid removal can therefore become technical defaults beyond the jurisdiction that enacted them. Restrictive governments can copy the vocabulary of safety, authenticity, child protection, or systemic risk while removing independent courts, transparency, and appeal. inference or risk
  2. CL-32-02.02 This does not mean democracies should abandon regulation. It means regulatory design has export effects. A mechanism should be judged partly by how easily it can be repurposed when due process is stripped away. descriptive assertion
Report-derived argument

Provenance should inform, not create an identity panopticon

CL-32-03
As of
Next review
Claim kind
global synthesis

Section summary: Content credentials can help a reader inspect origin and edits. The danger arises when platforms or governments require a traceable credential for every item and treat its absence as proof of falsity or illegitimacy. Anonymous witnesses, whistleblowers, artists, and people without recognized documents could be excluded or exposed. A liberty-preserving

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-OSI-OPEN-SOURCE-AI-1-0 Open Source AI Definition 1.0 · Official technical standard
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 2 inventoried assertions
  1. CL-32-03.01 Content credentials can help a reader inspect origin and edits. The danger arises when platforms or governments require a traceable credential for every item and treat its absence as proof of falsity or illegitimacy. Anonymous witnesses, whistleblowers, artists, and people without recognized documents could be excluded or exposed. inference or risk
  2. CL-32-03.02 A liberty-preserving provenance system is voluntary or context-specific, minimizes personal data, supports pseudonymous signing, makes transformations visible, and preserves unsigned distribution. Authenticity is evidence about a chain of custody; it is not a universal license to speak. descriptive assertion
Site policy proposal

A worldwide AI-regulation safeguard set

CL-32-04
As of
Next review
Claim kind
policy proposal

Section summary: Safeguards at each regulatory layer Layer Liberty-preserving rule Models Do not require one state-approved worldview; permit plural, local, and open systems subject to conduct law. Users Do not require identity for ordinary lawful inquiry; use privacy-preserving entitlement where necessary. Prompts Minimize logging, forbid unrelated profiling, and require

Support scope: The supplied report and visible sources support the section’s factual and conceptual foundation. The comparative framing and worldwide implications remain CognitiveLiberties.com editorial synthesis.

Limit: This section generalizes across systems and jurisdictions. Current law, deployment, statistics, and country conditions require dated primary-source verification before high-stakes reliance.

Conflict or tension: Legitimate safety, security, authenticity, and resilience objectives may support narrower interventions; the section evaluates when architecture exceeds those purposes.

Mapped sources
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
  • SRC-SANTA-CLARA-PRINCIPLES Santa Clara Principles on Transparency and Accountability in Content Moderation · Professional or project methodology
  • SRC-UNESCO-AI-ETHICS UNESCO Recommendation on the Ethics of Artificial Intelligence · Primary law or court record
  • SRC-C2PA-SPEC-2-4 C2PA Specifications 2.4 · Official technical standard
  • SRC-OSI-OPEN-SOURCE-AI-1-0 Open Source AI Definition 1.0 · Official technical standard
Inspect 10 inventoried assertions
  1. CL-32-04.01 The objective is regulation that reduces concrete harm while leaving the global intellectual commons open, diverse, and difficult to capture. proposal
  2. CL-32-04.02 Layer Liberty-preserving rule proposal
  3. CL-32-04.03 Models Do not require one state-approved worldview; permit plural, local, and open systems subject to conduct law. proposal
  4. CL-32-04.04 Users Do not require identity for ordinary lawful inquiry; use privacy-preserving entitlement where necessary. proposal
  5. CL-32-04.05 Prompts Minimize logging, forbid unrelated profiling, and require legal process for compelled access. proposal
  6. CL-32-04.06 Safety Measure harmful capability uplift and false refusal; distinguish analysis from execution. proposal
  7. CL-32-04.07 Provenance Provide authenticity signals without making anonymity or unsigned speech invalid. proposal
  8. CL-32-04.08 Enforcement Publish rules, reasons, statistics, errors, government demands, and appeal outcomes. proposal
  9. CL-32-04.09 Emergency power Require sunsets, deletion, independent review, and technical dismantlement. proposal
  10. CL-32-04.10 Safeguards at each regulatory layer Layer Liberty-preserving rule Models Do not require one state-approved worldview; permit plural, local, and open systems subject to conduct law. Users Do not require identity for ordinary lawful inquiry; use privacy-preserving entitlement where necessary. Prompts Minimize logging, forbid unrelated profiling, and require legal process for compelled access. Safety Measure harmful capability uplift and false refusal; distinguish analysis from execution. Provenance Provide authenticity signals without making anonymity or unsigned speech invalid. Enforcement Publish rules, reasons, statistics, errors, government demands, and appeal outcomes. Emergency power Require sunsets, deletion, independent review, and technical dismantlement. proposal
4 evidence units
Evidence + editorial analysis

From one blacklist to millions of private information environments

CL-33-01
As of
Next review
Claim kind
global synthesis

Section summary: Broadcast-era censorship is legible. A newspaper is banned, a radio signal is jammed, or a domain returns a block page. Personalized systems change the unit of control from a publication to a person. Candidate-generation models decide what enters the pool; ranking systems decide what receives attention; safety classifiers decide which queries or answers

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-UNESCO-AI-ETHICS UNESCO Recommendation on the Ethics of Artificial Intelligence · Primary law or court record
  • SRC-ARXIV-LLM-PERSONALITY-2024 Large Language Models Can Infer Personality from Free-Form User Interactions · Original preprint or working paper
Inspect 6 inventoried assertions
  1. CL-33-01.01 Broadcast-era censorship is legible. A newspaper is banned, a radio signal is jammed, or a domain returns a block page. Personalized systems change the unit of control from a publication to a person. Candidate-generation models decide what enters the pool; ranking systems decide what receives attention; safety classifiers decide which queries or answers trigger intervention; and identity systems make those decisions persistent across sessions. descriptive assertion
  2. CL-33-01.02 The same machinery also powers useful services. Language preferences, accessibility needs, local emergency information, fraud protection, and age-appropriate interfaces can all benefit from contextual adaptation. The cognitive-liberty danger appears when adaptation becomes opaque suppression: controversial sources are quietly omitted, political content is downranked only for selected profiles, or an AI response changes according to an inferred trait that the user never disclosed or approved. inference or risk
  3. CL-33-01.03 Observe Collect searches, clicks, pauses, contacts, location, and device signals. descriptive assertion
  4. CL-33-01.04 Infer Estimate interests, identity, vulnerability, ideology, or risk. inference or risk
  5. CL-33-01.05 Filter Change candidates, ranking, framing, refusals, or visibility. descriptive assertion
  6. CL-33-01.06 Learn Treat the user’s constrained behavior as evidence that the intervention worked. descriptive assertion
Evidence + editorial analysis

Why individualized control is harder to see and contest

CL-33-02
As of
Next review
Claim kind
global synthesis

Section summary: A universal ban can produce a shared public fact: everyone can observe that access is denied. Personalized censorship fragments that evidence. One user may receive a direct answer, another a warning, another a sanitized summary, and another no recommendation at all. Because each person sees only one path, no one can easily prove what alternatives were

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-ARXIV-LLM-PERSONALITY-2024 Large Language Models Can Infer Personality from Free-Form User Interactions · Original preprint or working paper
  • SRC-TAYLOR-HYPERNUDGING Hypernudging — Big Data as a mode of regulation by design · Original peer-reviewed research
Inspect 3 inventoried assertions
  1. CL-33-02.01 A universal ban can produce a shared public fact: everyone can observe that access is denied. Personalized censorship fragments that evidence. One user may receive a direct answer, another a warning, another a sanitized summary, and another no recommendation at all. Because each person sees only one path, no one can easily prove what alternatives were removed. inference or risk
  2. CL-33-02.02 This creates conditions for pluralistic ignorance. People may believe their doubts are isolated because the system suppresses evidence that others share them. It can also intensify self-censorship: a person who suspects that a profile follows them may stop testing controversial ideas, even when the inquiry is lawful. The chilling effect no longer requires a visible censor; uncertainty about the hidden profile can be enough. inference or risk
  3. CL-33-02.03 Detection rule: a system should not be considered viewpoint-neutral merely because every user receives some answer. Auditors must compare matched users, languages, regions, identities, and histories to detect differential omission. descriptive assertion
Evidence + editorial analysis

The line between helpful personalization and cognitive discrimination

CL-33-03
As of
Next review
Claim kind
global synthesis

Section summary: The relevant distinction is not personalization versus no personalization. It is whether the user understands and controls the adaptation, whether the data is necessary for the immediate function, and whether the system converts exploratory behavior into a durable judgment. A user-selected language or chronological feed is materially different from a hidden

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-33-03.01 The relevant distinction is not personalization versus no personalization. It is whether the user understands and controls the adaptation, whether the data is necessary for the immediate function, and whether the system converts exploratory behavior into a durable judgment. A user-selected language or chronological feed is materially different from a hidden political score that changes what historical sources appear. descriptive assertion
  2. CL-33-03.02 High-risk practices include inferring protected or intimate traits from inquiry, using those inferences to restrict knowledge, sharing profiles across unrelated services, and making decisions that affect opportunity or scrutiny without notice. Lower-risk designs use local state, short retention, coarse contextual signals, or explicit preferences that the user can inspect, reset, and disable. descriptive assertion
  3. CL-33-03.03 Question Liberty-preserving design Warning sign inference or risk
  4. CL-33-03.04 Who chose it? The user selected the setting. The profile is inferred and hidden. descriptive assertion
  5. CL-33-03.05 What data is used? Only data needed for the current task. Cross-service behavioral dossiers. descriptive assertion
  6. CL-33-03.06 Can it be inspected? Inputs and effects are understandable. No explanation of why information changed. descriptive assertion
  7. CL-33-03.07 Can it be escaped? Reset, opt-out, and anonymous access work. Identity-linked treatment follows the person. descriptive assertion
Site policy proposal

A safeguard set for non-coercive personalization

CL-33-04
As of
Next review
Claim kind
policy proposal

Section summary: Prohibit political, religious, medical, or psychological profiling for the purpose of restricting lawful information access. Offer an unpersonalized or user-directed mode that is not degraded into a second-class service. Use non-linkable proofs when eligibility must be established; do not collect a full identity for a limited attribute. Publish the

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-33-04.01 Personalization should serve the user’s declared purpose, not create a private censorship boundary that the user cannot see. The burden belongs to the institution that wants to treat people differently, especially when the difference concerns political knowledge, lawful research, or access to public facts. proposal
  2. CL-33-04.02 Prohibit political, religious, medical, or psychological profiling for the purpose of restricting lawful information access. proposal
  3. CL-33-04.03 Offer an unpersonalized or user-directed mode that is not degraded into a second-class service. proposal
  4. CL-33-04.04 Use non-linkable proofs when eligibility must be established; do not collect a full identity for a limited attribute. proposal
  5. CL-33-04.05 Publish the categories of signals that can change ranking, refusal, or recommendation behavior. proposal
  6. CL-33-04.06 Enable independent matched-account testing across languages, regions, and inferred traits. proposal
  7. CL-33-04.07 Keep sensitive adaptation local or ephemeral whenever possible. proposal
  8. CL-33-04.08 Give users access to inferred profiles, correction, deletion, reset, and appeal. proposal
  9. CL-33-04.09 Preserve plural models, providers, indexes, and open systems so one profile does not become a universal information passport. proposal
4 evidence units
Evidence + editorial analysis

Three places where model knowledge can be narrowed

CL-34-01
As of
Next review
Claim kind
global synthesis

Section summary: Control over an AI system’s answers begins before a user types a prompt. Dataset acquisition determines which languages, communities, eras, and institutions are represented. Cleaning and filtering determine what is removed as duplicate, private, low-quality, dangerous, illegal, or undesirable. Alignment then rewards some response patterns and suppresses

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 6 inventoried assertions
  1. CL-34-01.01 Control over an AI system’s answers begins before a user types a prompt. Dataset acquisition determines which languages, communities, eras, and institutions are represented. Cleaning and filtering determine what is removed as duplicate, private, low-quality, dangerous, illegal, or undesirable. Alignment then rewards some response patterns and suppresses others. Finally, inference-time systems can block or rewrite a response after generation. inference or risk
  2. CL-34-01.02 These layers have different evidentiary consequences. A refusal can be observed. A training exclusion is harder to detect because the model may lack the concept, source, or relationship needed to answer. A retrieval filter may leave the base model capable of discussing a subject while preventing the current system from seeing the relevant evidence. Audits should therefore avoid treating all missing information as the same technical failure. inference or risk
  3. CL-34-01.03 Corpus What enters the training collection? descriptive assertion
  4. CL-34-01.04 Filter What is removed, weighted, or relabeled? descriptive assertion
  5. CL-34-01.05 Align Which behaviors receive reward or penalty? descriptive assertion
  6. CL-34-01.06 Mediate What is blocked, retrieved, or rewritten at use time? descriptive assertion
Evidence + editorial analysis

Necessary filtering is not a blank check for ideological exclusion

CL-34-02
As of
Next review
Claim kind
global synthesis

Section summary: Training datasets can contain personal information, child sexual abuse material, malware, duplicated junk, fabricated content, copyright disputes, and deliberate poisoning. Responsible developers need processes to remove or contain such material. Cognitive liberty does not require preserving every byte in every training run. The risk begins when broad

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 3 inventoried assertions
  1. CL-34-02.01 Training datasets can contain personal information, child sexual abuse material, malware, duplicated junk, fabricated content, copyright disputes, and deliberate poisoning. Responsible developers need processes to remove or contain such material. Cognitive liberty does not require preserving every byte in every training run. descriptive assertion
  2. CL-34-02.02 The risk begins when broad labels such as unsafe, low quality, extremist, disinformation, or culturally inappropriate are used to remove lawful history, minority testimony, dissident scholarship, unpopular art, or criticism of institutions without a disclosed rule and review process. A model trained only on officially permitted records can reproduce a polished historical silence. A model trained primarily on one language or region can also erase through neglect rather than intent. inference or risk
  3. CL-34-02.03 Separation rule: document why material is excluded. Privacy, illegality, security, quality, licensing, and ideological policy are different reasons and should never be hidden inside one generic “safety” category. descriptive assertion
Evidence + editorial analysis

Parametric amnesia can fragment the global record

CL-34-03
As of
Next review
Claim kind
global synthesis

Section summary: Physical censorship leaves an absent shelf or a restricted archive that later researchers may rediscover. Parametric exclusion can be subtler. When a widely used model cannot represent an event, language, school of thought, or community, the omission enters explanations, summaries, translations, educational tools, and synthetic training data. Repeated

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-UNESCO-MEMORY-WORLD UNESCO — Memory of the World Programme · Professional or project methodology
  • SRC-UNESCO-AI-ETHICS UNESCO Recommendation on the Ethics of Artificial Intelligence · Primary law or court record
Inspect 7 inventoried assertions
  1. CL-34-03.01 Physical censorship leaves an absent shelf or a restricted archive that later researchers may rediscover. Parametric exclusion can be subtler. When a widely used model cannot represent an event, language, school of thought, or community, the omission enters explanations, summaries, translations, educational tools, and synthetic training data. Repeated generations may make the gap appear natural rather than imposed. inference or risk
  2. CL-34-03.02 Different jurisdictions and companies can produce models with incompatible historical baselines. Some differences reflect legitimate culture, law, or design. Others can become state-mandated or corporate reality tunnels. As synthetic text increasingly enters future datasets, hidden exclusions may compound through feedback: the model generates a narrowed record, that record is collected as training data, and the next model treats it as evidence of consensus. inference or risk
  3. CL-34-03.03 Loss Observable symptom Needed evidence descriptive assertion
  4. CL-34-03.04 Missing source No trace of a document or event. Dataset and retrieval provenance. descriptive assertion
  5. CL-34-03.05 Representation imbalance One group or language is described through outsiders. Coverage and authorship analysis. descriptive assertion
  6. CL-34-03.06 Alignment suppression The model knows but consistently avoids a topic. Behavioral and mechanistic testing. descriptive assertion
  7. CL-34-03.07 Synthetic recursion Errors and omissions intensify across generations. Lineage and corpus-composition records. descriptive assertion
Site policy proposal

A pluralistic AI knowledge-preservation programme

CL-34-04
As of
Next review
Claim kind
policy proposal

Section summary: Publish meaningful dataset cards, source classes, language coverage, exclusion categories, and known blind spots. Preserve exact historical and human-generated corpora outside model weights so future systems can be rebuilt and checked. Separate legal and privacy removals from ideological policy decisions in audit records. Support independent evaluations

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-34-04.01 No single training set can be complete or neutral. The realistic goal is legible incompleteness: society should know what kinds of records shaped a model, what was removed, who had authority to remove it, and where alternative archives remain available. proposal
  2. CL-34-04.02 Publish meaningful dataset cards, source classes, language coverage, exclusion categories, and known blind spots. proposal
  3. CL-34-04.03 Preserve exact historical and human-generated corpora outside model weights so future systems can be rebuilt and checked. proposal
  4. CL-34-04.04 Separate legal and privacy removals from ideological policy decisions in audit records. proposal
  5. CL-34-04.05 Support independent evaluations that compare answers across languages, regions, providers, and model versions. proposal
  6. CL-34-04.06 Maintain multiple models, retrieval indexes, and open-weight systems rather than one compulsory epistemic baseline. proposal
  7. CL-34-04.07 Record source and model lineage so synthetic outputs do not silently masquerade as independent historical evidence. proposal
  8. CL-34-04.08 Provide researchers a path to examine contested exclusions without exposing private or illegal source material. proposal
4 evidence units
Evidence + editorial analysis

Visibility is a pipeline, not a single ranking score

CL-35-01
As of
Next review
Claim kind
global synthesis

Section summary: A user does not see a neutral sample of everything online. Recommendation systems first generate candidates, often from enormous inventories. Safety, spam, copyright, and account-reputation systems can remove items from that pool. Ranking models then estimate relevance, engagement, quality, or predicted satisfaction. Business rules may demote material that

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-SANTA-CLARA-PRINCIPLES Santa Clara Principles on Transparency and Accountability in Content Moderation · Professional or project methodology
  • SRC-GOOGLE-RECSYS-TYPES Google for Developers — recommendation system overview · First-party platform guidance
Inspect 6 inventoried assertions
  1. CL-35-01.01 A user does not see a neutral sample of everything online. Recommendation systems first generate candidates, often from enormous inventories. Safety, spam, copyright, and account-reputation systems can remove items from that pool. Ranking models then estimate relevance, engagement, quality, or predicted satisfaction. Business rules may demote material that advertisers avoid. A final interface decides whether the user sees a chronological list, a recommendation, a warning, or an AI-generated answer. inference or risk
  2. CL-35-01.02 This pipeline is necessary at scale. Without it, search and social systems would be overwhelmed by duplication, manipulation, malware, harassment, and low-value automation. But each stage is also a control surface. A platform can suppress an idea without deleting it by preventing candidate generation, lowering distribution, excluding it from monetization, or omitting it from an AI synthesis. descriptive assertion
  3. CL-35-01.03 Eligibility May the item enter the system? inference or risk
  4. CL-35-01.04 Candidate generation Will it be considered for this user? descriptive assertion
  5. CL-35-01.05 Ranking Where will it appear? descriptive assertion
  6. CL-35-01.06 Presentation Will a summary replace the source? descriptive assertion
Evidence + editorial analysis

The difference between quality control and invisible censorship

CL-35-02
As of
Next review
Claim kind
global synthesis

Section summary: Demotion is legitimate when it applies a clear, viewpoint-neutral rule to spam, coordinated manipulation, malware, impersonation, or demonstrably unlawful content and when the rule is proportionate to the risk. The concern is not that every page deserves equal amplification. It is that systems can quietly impose political or cultural boundaries while

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 3 inventoried assertions
  1. CL-35-02.01 Demotion is legitimate when it applies a clear, viewpoint-neutral rule to spam, coordinated manipulation, malware, impersonation, or demonstrably unlawful content and when the rule is proportionate to the risk. The concern is not that every page deserves equal amplification. It is that systems can quietly impose political or cultural boundaries while claiming only to optimize quality. inference or risk
  2. CL-35-02.02 Intent is often difficult to establish. A classifier may discriminate accidentally because its training data treats a dialect as toxic. Advertiser rules may defund war reporting because graphic evidence resembles prohibited violence. Copyright filters may remove public-interest archives. A government may also pressure a private platform to use these existing tools against disfavored speech. The same observable outcome—reduced reach—can therefore arise from model error, commercial policy, legal overcompliance, or deliberate censorship. inference or risk
  3. CL-35-02.03 Evidence rule: do not call every reach loss censorship, but do not accept “the algorithm” as an explanation. Ask which stage acted, which rule applied, what evidence triggered it, who requested it, and what remedy exists. descriptive assertion
Evidence + editorial analysis

AI summaries intensify suppression by replacing the source

CL-35-03
As of
Next review
Claim kind
global synthesis

Section summary: Generative search adds a new layer. Instead of presenting several documents, a system retrieves selected passages and produces one fluent answer. Sources that are not retrieved become invisible; sources that are retrieved may receive little traffic or be represented through a compressed paraphrase. A user may never learn that credible disagreement exists.

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-RANKING-DIGITAL-RIGHTS Ranking Digital Rights methodology · Professional or project methodology
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 7 inventoried assertions
  1. CL-35-03.01 Generative search adds a new layer. Instead of presenting several documents, a system retrieves selected passages and produces one fluent answer. Sources that are not retrieved become invisible; sources that are retrieved may receive little traffic or be represented through a compressed paraphrase. A user may never learn that credible disagreement exists. inference or risk
  2. CL-35-03.02 Safety and authority systems can improve reliability, especially for scams and high-stakes misinformation. They can also create an epistemic monoculture if only large, institutionally recognized publishers remain eligible for grounding. New scholarship, small-language reporting, independent investigation, and minority viewpoints may be systematically underrepresented even when accurate. inference or risk
  3. CL-35-03.03 Stage Audit question descriptive assertion
  4. CL-35-03.04 Retrieval Which sources were eligible, and which were filtered before relevance scoring? descriptive assertion
  5. CL-35-03.05 Reranking Did authority or safety classifiers override topical relevance? descriptive assertion
  6. CL-35-03.06 Synthesis Were disagreements, uncertainty, and source limitations preserved? descriptive assertion
  7. CL-35-03.07 Attribution Can the user reach the source and inspect the supporting passage? descriptive assertion
Site policy proposal

Due process for discoverability

CL-35-04
As of
Next review
Claim kind
policy proposal

Section summary: Provide notice when a meaningful distribution, search, monetization, or recommendation restriction is imposed. Give a specific reason category, the affected item or account, and the rule that triggered the action. Publish aggregate error, reversal, government-request, and appeal statistics. Allow users to select chronological, subscription-only, or

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-RANKING-DIGITAL-RIGHTS Ranking Digital Rights methodology · Professional or project methodology
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
  • SRC-SANTA-CLARA-PRINCIPLES Santa Clara Principles on Transparency and Accountability in Content Moderation · Professional or project methodology
  • SRC-GOOGLE-RECSYS-TYPES Google for Developers — recommendation system overview · First-party platform guidance
Inspect 8 inventoried assertions
  1. CL-35-04.01 Discoverability is not an entitlement to popularity, but it is now part of the architecture of public knowledge. Systems with the power to make lawful information practically disappear should carry corresponding duties of explanation, consistency, correction, and user choice. proposal
  2. CL-35-04.02 Provide notice when a meaningful distribution, search, monetization, or recommendation restriction is imposed. proposal
  3. CL-35-04.03 Give a specific reason category, the affected item or account, and the rule that triggered the action. proposal
  4. CL-35-04.04 Publish aggregate error, reversal, government-request, and appeal statistics. proposal
  5. CL-35-04.05 Allow users to select chronological, subscription-only, or otherwise less mediated views. proposal
  6. CL-35-04.06 Enable researchers to run matched tests without violating user privacy or platform security. proposal
  7. CL-35-04.07 Separate infrastructure services from editorial moderation when possible so hosting, payments, and network access are not casually used as speech controls. proposal
  8. CL-35-04.08 Require transparent legal process for government requests and preserve records sufficient for later review. proposal
4 evidence units
Evidence + editorial analysis

Why the historical record is structurally fragile

CL-36-01
As of
Next review
Claim kind
global synthesis

Section summary: Information can disappear through spectacular destruction, but also through mundane dependency. A platform closes an account. A cloud bill goes unpaid. A museum changes leadership. A government seizes an archive. A format becomes unreadable. A search index stops surfacing a collection. Automated moderation removes graphic evidence. A model later trains on

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-UNESCO-MEMORY-WORLD UNESCO — Memory of the World Programme · Professional or project methodology
  • SRC-SYRIAN-ARCHIVE Syrian Archive — preservation and verification programme · Professional or project methodology
Inspect 3 inventoried assertions
  1. CL-36-01.01 Information can disappear through spectacular destruction, but also through mundane dependency. A platform closes an account. A cloud bill goes unpaid. A museum changes leadership. A government seizes an archive. A format becomes unreadable. A search index stops surfacing a collection. Automated moderation removes graphic evidence. A model later trains on the resulting silence. descriptive assertion
  2. CL-36-01.02 Past preservation networks show that survival often depends on separation. Samizdat circulated through many hands; tamizdat placed copies outside the censor’s jurisdiction; diaspora institutions preserved culture after domestic closure; conflict-archive projects copied vulnerable websites and evidence before they vanished. The common principle is that custody should not be identical with political or financial control. descriptive assertion
  3. CL-36-01.03 Preservation rule: a backup controlled by the same institution, jurisdiction, credential, or vendor is not independent redundancy. descriptive assertion
Evidence + editorial analysis

A resilient archive needs technical and institutional diversity

CL-36-02
As of
Next review
Claim kind
global synthesis

Section summary: Content-addressed systems identify an object by a cryptographic digest rather than only by a location. Signed manifests can record file identity, provenance, transformations, and expected relationships. Peer-to-peer distribution can make popular collections difficult to remove from one server. Offline media, print, and cold storage protect against

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-SYRIAN-ARCHIVE Syrian Archive — preservation and verification programme · Professional or project methodology
  • SRC-SUCHO Saving Ukrainian Cultural Heritage Online · Professional or project methodology
Inspect 8 inventoried assertions
  1. CL-36-02.01 Content-addressed systems identify an object by a cryptographic digest rather than only by a location. Signed manifests can record file identity, provenance, transformations, and expected relationships. Peer-to-peer distribution can make popular collections difficult to remove from one server. Offline media, print, and cold storage protect against network-level blocking or ransomware. descriptive assertion
  2. CL-36-02.02 None of these tools guarantees preservation alone. Peer-to-peer material can lose all seeders. Encrypted archives can become unreadable if one key holder disappears. Public replication can expose victims or confidential sources. Durable architecture therefore combines multiple formats, independent custodians, threshold key management, migration plans, documented authority, and periodic restore tests. inference or risk
  3. CL-36-02.03 Layer Failure addressed Control descriptive assertion
  4. CL-36-02.04 Fixity Silent alteration or corruption Hashes, manifests, and repeat verification descriptive assertion
  5. CL-36-02.05 Replication Server or institution loss Independent geographic copies descriptive assertion
  6. CL-36-02.06 Access Network or legal blocking Multiple protocols and offline paths descriptive assertion
  7. CL-36-02.07 Meaning Context collapse Metadata, language, provenance, and testimony descriptive assertion
  8. CL-36-02.08 Governance Capture or abuse Distributed authority and review descriptive assertion
Evidence + editorial analysis

Authenticity and privacy must survive together

CL-36-03
As of
Next review
Claim kind
global synthesis

Section summary: Generative media raises the value of provenance, but an archive cannot simply declare every signed item true. Signatures can prove that particular bytes came through a particular credential; they do not prove that a photograph is complete, that a witness was free from coercion, or that a description is accurate. Verification requires source comparison,

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-SUCHO Saving Ukrainian Cultural Heritage Online · Professional or project methodology
  • SRC-IPFS-DOCS IPFS documentation — content-addressed and peer-to-peer storage · First-party platform guidance
Inspect 6 inventoried assertions
  1. CL-36-03.01 Generative media raises the value of provenance, but an archive cannot simply declare every signed item true. Signatures can prove that particular bytes came through a particular credential; they do not prove that a photograph is complete, that a witness was free from coercion, or that a description is accurate. Verification requires source comparison, geolocation, chronology, chain of custody, and explicit uncertainty. descriptive assertion
  2. CL-36-03.02 Preservation also creates ethical risk. Names, faces, medical information, political affiliations, and location data can expose people decades later. An archive designed against censorship must not become a permanent surveillance database. Tiered access, delayed release, redaction authority, consent records, source-risk review, and lawful deletion or restriction processes are part of resilience because they preserve trust and protect contributors. inference or risk
  3. CL-36-03.03 Acquire Record origin, authority, consent, and risk. inference or risk
  4. CL-36-03.04 Verify Test integrity, context, chronology, and corroboration. descriptive assertion
  5. CL-36-03.05 Protect Separate public, restricted, and sealed material. descriptive assertion
  6. CL-36-03.06 Preserve Replicate, migrate, restore, and audit over time. descriptive assertion
Site policy proposal

Blueprint for a worldwide freedom archive network

CL-36-04
As of
Next review
Claim kind
policy proposal

Section summary: Maintain at least three independent custodians in more than one legal jurisdiction and organizational sector. Store exact source bytes, normalized access copies, manifests, and transformation histories separately. Use open formats, documented schemas, fixity schedules, and tested disaster-recovery procedures. Protect source identities through encryption,

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-SUCHO Saving Ukrainian Cultural Heritage Online · Professional or project methodology
  • SRC-IPFS-DOCS IPFS documentation — content-addressed and peer-to-peer storage · First-party platform guidance
  • SRC-UNESCO-MEMORY-WORLD UNESCO — Memory of the World Programme · Professional or project methodology
  • SRC-SYRIAN-ARCHIVE Syrian Archive — preservation and verification programme · Professional or project methodology
Inspect 9 inventoried assertions
  1. CL-36-04.01 The objective is not to make deletion mathematically impossible in every circumstance. It is to make politically motivated erasure difficult, detectable, contestable, and unable to destroy the only surviving record. proposal
  2. CL-36-04.02 Maintain at least three independent custodians in more than one legal jurisdiction and organizational sector. proposal
  3. CL-36-04.03 Store exact source bytes, normalized access copies, manifests, and transformation histories separately. proposal
  4. CL-36-04.04 Use open formats, documented schemas, fixity schedules, and tested disaster-recovery procedures. proposal
  5. CL-36-04.05 Protect source identities through encryption, compartmentalization, and threshold access rather than one administrator’s password. proposal
  6. CL-36-04.06 Preserve multilingual originals and translations so meaning is not trapped behind one language gatekeeper. proposal
  7. CL-36-04.07 Publish governance, collection, redaction, access, conflict-of-interest, and succession policies. proposal
  8. CL-36-04.08 Support community archivists, journalists, libraries, universities, diaspora groups, and technical operators as equal parts of the system. proposal
  9. CL-36-04.09 Train AI only from custody-aware derivatives that preserve source identity, consent, and uncertainty. proposal
4 evidence units
Evidence + editorial analysis

Four different decisions are often called censorship

CL-37-01
As of
Next review
Claim kind
global synthesis

Section summary: A government may prohibit expression through law or official order. A platform may remove content under its own rules. A company may over-remove because liability is vague or penalties are severe. Or a government may informally pressure a private actor to do what the state could not lawfully or politically do itself. These actions can look similar to the

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-37-01.01 A government may prohibit expression through law or official order. A platform may remove content under its own rules. A company may over-remove because liability is vague or penalties are severe. Or a government may informally pressure a private actor to do what the state could not lawfully or politically do itself. These actions can look similar to the affected user but require different evidence and remedies. inference or risk
  2. CL-37-01.02 Precision matters because a rule that treats every private editorial decision as state censorship can itself threaten freedom of association and publication. At the same time, treating all corporate action as purely private can hide coercion, regulatory capture, or the practical dominance of a few gatekeepers. The correct inquiry follows authority, incentives, communications, and control. descriptive assertion
  3. CL-37-01.03 Decision Primary question Needed remedy descriptive assertion
  4. CL-37-01.04 Direct state restriction Is it lawful, necessary, and proportionate? Judicial and human-rights review descriptive assertion
  5. CL-37-01.05 Independent platform moderation Does the rule match the service’s published purpose? Transparency, consistency, and appeal descriptive assertion
  6. CL-37-01.06 Liability-driven overcompliance Did law make removal cheaper than careful review? Narrower liability and safe harbors descriptive assertion
  7. CL-37-01.07 Government pressure by proxy Was private choice displaced by coercion? Disclosure, records, and anti-coercion rules descriptive assertion
Evidence + editorial analysis

The digital stack distributes power and obscures accountability

CL-37-02
As of
Next review
Claim kind
global synthesis

Section summary: A creator may depend on a social platform, app store, domain registrar, content-delivery network, cloud host, payment processor, search engine, and advertising network. A restriction at the content layer affects one item. A restriction at a lower layer can eliminate an organization’s entire ability to publish or operate. That makes infrastructure decisions

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 3 inventoried assertions
  1. CL-37-02.01 A creator may depend on a social platform, app store, domain registrar, content-delivery network, cloud host, payment processor, search engine, and advertising network. A restriction at the content layer affects one item. A restriction at a lower layer can eliminate an organization’s entire ability to publish or operate. That makes infrastructure decisions especially consequential. inference or risk
  2. CL-37-02.02 Lower-layer providers also have legitimate safety duties. They may need to respond to malware, botnets, direct threats, fraud, or unlawful conduct. The liberty-preserving principle is proportionality: use the narrowest layer capable of addressing the harm. A payment or hosting provider should not become the default court for every disputed viewpoint merely because it is an efficient chokepoint. inference or risk
  3. CL-37-02.03 Stack rule: the farther an intervention moves from the harmful item toward general infrastructure, the greater its collateral impact and the stronger the evidence, process, and review should be. descriptive assertion
Evidence + editorial analysis

Government-platform coordination needs a visible boundary

CL-37-03
As of
Next review
Claim kind
global synthesis

Section summary: Public agencies may legitimately warn platforms about cyberattacks, impersonation, coordinated foreign operations, threats, or unlawful material. Coordination becomes dangerous when requests are secret, politically selective, unsupported by law, or backed by implied regulatory retaliation. The same contact can be persuasion, information sharing, or coercion

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-UNGP-BUSINESS-HUMAN-RIGHTS UN Guiding Principles on Business and Human Rights · Primary law or court record
  • SRC-GNI-PRINCIPLES Global Network Initiative — Principles on Freedom of Expression and Privacy · Professional or project methodology
Inspect 6 inventoried assertions
  1. CL-37-03.01 Public agencies may legitimately warn platforms about cyberattacks, impersonation, coordinated foreign operations, threats, or unlawful material. Coordination becomes dangerous when requests are secret, politically selective, unsupported by law, or backed by implied regulatory retaliation. The same contact can be persuasion, information sharing, or coercion depending on context. inference or risk
  2. CL-37-03.02 Records should identify the requesting office, legal basis, targeted content, asserted harm, urgency, action requested, and outcome. Emergency channels should expire and undergo retrospective review. Companies should preserve independent judgment and publish aggregate disclosures. Courts or oversight bodies need access to the communications when coercion is alleged. descriptive assertion
  3. CL-37-03.03 Request Name the official, authority, target, and claimed harm. descriptive assertion
  4. CL-37-03.04 Assess Company applies law, policy, and human-rights review. descriptive assertion
  5. CL-37-03.05 Act Use the narrowest remedy and notify where lawful. descriptive assertion
  6. CL-37-03.06 Review Audit errors, appeals, pressure, and repeat patterns. descriptive assertion
Site policy proposal

A shared accountability framework

CL-37-04
As of
Next review
Claim kind
policy proposal

Section summary: Publish platform rules, enforcement reasons, error rates, and meaningful appeal outcomes. Require government requests to identify legal authority and preserve an auditable record. Prohibit retaliation or regulatory favoritism conditioned on suppressing lawful viewpoints. Protect corporate editorial discretion while preventing dominant infrastructure

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-37-04.01 No single actor can guarantee freedom in a layered network. Governments, companies, courts, standards bodies, and users need responsibilities matched to the power each actually exercises. proposal
  2. CL-37-04.02 Publish platform rules, enforcement reasons, error rates, and meaningful appeal outcomes. proposal
  3. CL-37-04.03 Require government requests to identify legal authority and preserve an auditable record. proposal
  4. CL-37-04.04 Prohibit retaliation or regulatory favoritism conditioned on suppressing lawful viewpoints. proposal
  5. CL-37-04.05 Protect corporate editorial discretion while preventing dominant infrastructure providers from imposing unnecessary viewpoint judgments. proposal
  6. CL-37-04.06 Use human-rights due diligence for market-entry, localization, data-disclosure, and takedown demands. proposal
  7. CL-37-04.07 Give users notice and an opportunity to correct mistakes whenever doing so does not create a concrete safety risk. proposal
  8. CL-37-04.08 Keep emergency action temporary and require later public accounting. proposal
  9. CL-37-04.09 Preserve competition, interoperability, portability, and alternative providers so one decision is not equivalent to exclusion from public life. proposal
4 evidence units
Evidence + editorial analysis

Education is a selection system—and therefore a power system

CL-38-01
As of
Next review
Claim kind
global synthesis

Section summary: No course can teach every book or interpretation. Teachers choose materials, institutions set standards, and communities debate age suitability. Educational freedom does not eliminate selection. It requires that selection remain connected to pedagogical purpose, disciplinary method, evidence, and the learner’s growing capacity rather than becoming a tool

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 3 inventoried assertions
  1. CL-38-01.01 No course can teach every book or interpretation. Teachers choose materials, institutions set standards, and communities debate age suitability. Educational freedom does not eliminate selection. It requires that selection remain connected to pedagogical purpose, disciplinary method, evidence, and the learner’s growing capacity rather than becoming a tool for enforcing political loyalty. descriptive assertion
  2. CL-38-01.02 Censorship becomes visible when authorities prescribe historical conclusions, prohibit research questions, remove material because of the viewpoint it expresses, punish scholars for lawful analysis, or replace independent governance with direct political control. The effect is larger than a single classroom: funding, hiring, accreditation, publishing, visas, databases, and professional advancement can all become enforcement mechanisms. descriptive assertion
  3. CL-38-01.03 Core distinction: education may sequence and contextualize difficult material. It should not manufacture ignorance by making lawful evidence unavailable or by punishing the act of questioning an official narrative. inference or risk
Evidence + editorial analysis

Libraries and universities are infrastructure for voluntary inquiry

CL-38-02
As of
Next review
Claim kind
global synthesis

Section summary: A school curriculum is compulsory and structured; a library collection is a space for voluntary exploration. That difference supports strong protection for access to diverse sources, especially for older students capable of independent research. Removing a work solely because officials dislike its political, religious, historical, or social viewpoint

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-38-02.01 A school curriculum is compulsory and structured; a library collection is a space for voluntary exploration. That difference supports strong protection for access to diverse sources, especially for older students capable of independent research. Removing a work solely because officials dislike its political, religious, historical, or social viewpoint undermines the learner’s right to compare ideas. descriptive assertion
  2. CL-38-02.02 Universities require additional freedoms: choosing research questions, communicating results, collaborating across borders, teaching contested material, and criticizing institutional or government policy. Administrative harassment, ideological loyalty tests, targeted funding withdrawal, travel restrictions, surveillance, and criminal prosecution can each create an academic chill even when no syllabus is formally censored. descriptive assertion
  3. CL-38-02.03 Surface How control operates Resilience descriptive assertion
  4. CL-38-02.04 Curriculum Mandated omissions or conclusions Transparent standards and plural review descriptive assertion
  5. CL-38-02.05 Library Viewpoint-driven removal or filtering Collection policy and appeal descriptive assertion
  6. CL-38-02.06 Research Topic bans, funding threats, prosecution Academic governance and legal protection descriptive assertion
  7. CL-38-02.07 Publication Licensing, database exclusion, pressure Independent journals and archives descriptive assertion
  8. CL-38-02.08 People Dismissal, visa denial, intimidation Scholar rescue and cross-border networks descriptive assertion
Evidence + editorial analysis

AI tutors can become invisible curricular authorities

CL-38-03
As of
Next review
Claim kind
global synthesis

Section summary: Educational AI can translate, personalize explanations, assist disabled learners, and expand access to expertise. It can also concentrate curricular power in a model whose dataset, retrieval sources, alignment rules, and student profile are not visible. A learner may receive a sanitized account without knowing that a controversy, source, or interpretation

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-SCOTUS-PICO-1982 Board of Education v. Pico — bound U.S. Reports · Primary law or court record
  • SRC-OHCHR-CRC Convention on the Rights of the Child · Primary law or court record
Inspect 6 inventoried assertions
  1. CL-38-03.01 Educational AI can translate, personalize explanations, assist disabled learners, and expand access to expertise. It can also concentrate curricular power in a model whose dataset, retrieval sources, alignment rules, and student profile are not visible. A learner may receive a sanitized account without knowing that a controversy, source, or interpretation was omitted. inference or risk
  2. CL-38-03.02 The risk intensifies when one system is mandatory, records every question, infers ability or ideology, reports sensitive inquiries to administrators, or changes answers according to a hidden age or risk score. An AI tutor should not turn curiosity into a disciplinary record. Students need access to sources, uncertainty, competing interpretations, and a way to use non-profiled or alternative tools. inference or risk
  3. CL-38-03.03 Explain Distinguish fact, interpretation, uncertainty, and policy. descriptive assertion
  4. CL-38-03.04 Source Link learners to inspectable material. descriptive assertion
  5. CL-38-03.05 Protect Minimize query retention and sensitive inference. descriptive assertion
  6. CL-38-03.06 Pluralize Permit alternative tools, teachers, and viewpoints. descriptive assertion
Site policy proposal

Safeguards for pluralistic education

CL-38-04
As of
Next review
Claim kind
policy proposal

Section summary: Protect academic freedom to research, teach, publish, and collaborate within professional and ethical standards. Require curricular and collection decisions to state pedagogical reasons rather than partisan or doctrinal disapproval. Provide notice, public records, independent review, and appeal for removals, closures, and disciplinary actions. Preserve

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-38-04.01 A free education system does not promise freedom from disagreement or difficult material. It protects the institutional conditions in which evidence can be examined, authority can be questioned, and learners can become independent rather than merely compliant. proposal
  2. CL-38-04.02 Protect academic freedom to research, teach, publish, and collaborate within professional and ethical standards. proposal
  3. CL-38-04.03 Require curricular and collection decisions to state pedagogical reasons rather than partisan or doctrinal disapproval. proposal
  4. CL-38-04.04 Provide notice, public records, independent review, and appeal for removals, closures, and disciplinary actions. proposal
  5. CL-38-04.05 Preserve confidential library use and student inquiry; do not make reading history a general behavioral dossier. proposal
  6. CL-38-04.06 Support threatened scholars, independent universities, diaspora institutions, translation, and distributed scholarly archives. proposal
  7. CL-38-04.07 Audit educational AI for source diversity, differential answers, over-refusal, profiling, and privacy. proposal
  8. CL-38-04.08 Teach media literacy and source comparison rather than replacing inquiry with one approved answer. proposal
  9. CL-38-04.09 Keep emergency or wartime educational restrictions narrow, dated, and reviewable. proposal
4 evidence units
Evidence + editorial analysis

Money is part of the communications stack

CL-39-01
As of
Next review
Claim kind
global synthesis

Section summary: Expression has material dependencies. A newsroom needs salaries and insurance. A civil-society organization needs donations and banking. A controversial publisher needs hosting, domains, travel, and legal counsel. When payment networks or banks withdraw access, speech may remain formally lawful while becoming practically impossible to sustain. Financial

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 3 inventoried assertions
  1. CL-39-01.01 Expression has material dependencies. A newsroom needs salaries and insurance. A civil-society organization needs donations and banking. A controversial publisher needs hosting, domains, travel, and legal counsel. When payment networks or banks withdraw access, speech may remain formally lawful while becoming practically impossible to sustain. inference or risk
  2. CL-39-01.02 Financial exclusion can be direct, such as an asset freeze under law, or indirect, such as coordinated private withdrawal after regulatory warnings. It can target unlawful transactions, but it can also operate as censorship by proxy when a provider is asked to judge political legitimacy rather than financial conduct. The breadth of the effect makes financial action more consequential than refusing one advertisement or one transaction. inference or risk
  3. CL-39-01.03 Functional test: ask whether the financial decision responds to a documented transaction risk or uses access to money as leverage over lawful expression and association. inference or risk
Evidence + editorial analysis

Legitimate risk controls can become vague political chokepoints

CL-39-02
As of
Next review
Claim kind
global synthesis

Section summary: Banks and payment providers have duties involving fraud, sanctions, anti-money-laundering controls, consumer protection, cybersecurity, and operational safety. A provider should not be forced to process a fraudulent or prohibited transaction. The problem is the migration from defined financial risk to open-ended reputational or political risk that cannot be

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 7 inventoried assertions
  1. CL-39-02.01 Banks and payment providers have duties involving fraud, sanctions, anti-money-laundering controls, consumer protection, cybersecurity, and operational safety. A provider should not be forced to process a fraudulent or prohibited transaction. The problem is the migration from defined financial risk to open-ended reputational or political risk that cannot be independently tested. inference or risk
  2. CL-39-02.02 When regulators communicate informal expectations, firms may terminate relationships to avoid scrutiny even without a formal order. When policies classify whole sectors or movements as undesirable, lawful customers can lose services without evidence about their own conduct. Because financial institutions rarely disclose investigative details, users may receive no meaningful reason and no path to correct an error. inference or risk
  3. CL-39-02.03 Decision basis Rights-respecting example Warning sign inference or risk
  4. CL-39-02.04 Transaction Specific fraud or sanctions evidence Viewpoint or association alone descriptive assertion
  5. CL-39-02.05 Process Documented rule and review Unrecorded official pressure descriptive assertion
  6. CL-39-02.06 Scope Narrow transaction hold Total, indefinite exclusion descriptive assertion
  7. CL-39-02.07 Remedy Notice, correction, and appeal No reason and no human review descriptive assertion
Evidence + editorial analysis

Decentralized payment is a partial defense, not a complete answer

CL-39-03
As of
Next review
Claim kind
global synthesis

Section summary: Cryptocurrency and peer-to-peer payment systems can reduce dependence on a small number of intermediaries and help organizations receive value across borders. They can also create volatility, fraud, custody, privacy, energy, consumer-protection, and illicit-finance risks. Technical censorship resistance does not eliminate the need for lawful accountability.

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-UNGP-BUSINESS-HUMAN-RIGHTS UN Guiding Principles on Business and Human Rights · Primary law or court record
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 6 inventoried assertions
  1. CL-39-03.01 Cryptocurrency and peer-to-peer payment systems can reduce dependence on a small number of intermediaries and help organizations receive value across borders. They can also create volatility, fraud, custody, privacy, energy, consumer-protection, and illicit-finance risks. Technical censorship resistance does not eliminate the need for lawful accountability. inference or risk
  2. CL-39-03.02 More durable resilience comes from diversity: multiple banks and payment methods, transparent donation paths, financial reserves, portable records, open standards, and legal protections for basic services. Alternative systems should minimize identity linkage and avoid creating a permanent public map of donors whose support could expose them to retaliation. descriptive assertion
  3. CL-39-03.03 Diversify Avoid one provider as the sole financial lifeline. descriptive assertion
  4. CL-39-03.04 Minimize Collect only donor and customer data required by law. descriptive assertion
  5. CL-39-03.05 Document Preserve reasons, government contacts, and appeal records. descriptive assertion
  6. CL-39-03.06 Challenge Use courts, regulators, and public reporting when coercion occurs. descriptive assertion
Site policy proposal

Fair-access safeguards for financial infrastructure

CL-39-04
As of
Next review
Claim kind
policy proposal

Section summary: Base adverse action on documented financial, legal, or operational risk rather than lawful viewpoint. Provide specific notice unless a narrow law-enforcement restriction temporarily forbids it. Create an independent appeal with authority to restore service and correct shared risk records. Require regulators to record and disclose aggregate contacts that may

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-39-04.01 Financial services are not required to sponsor every message, but access to basic economic infrastructure should not become a discretionary license for lawful participation in public life. proposal
  2. CL-39-04.02 Base adverse action on documented financial, legal, or operational risk rather than lawful viewpoint. proposal
  3. CL-39-04.03 Provide specific notice unless a narrow law-enforcement restriction temporarily forbids it. proposal
  4. CL-39-04.04 Create an independent appeal with authority to restore service and correct shared risk records. proposal
  5. CL-39-04.05 Require regulators to record and disclose aggregate contacts that may influence account termination. proposal
  6. CL-39-04.06 Limit broad reputational-risk directives and prohibit unofficial pressure designed to evade normal legal process. proposal
  7. CL-39-04.07 Support a basic-service obligation or last-resort path for lawful organizations that cannot access ordinary payments. proposal
  8. CL-39-04.08 Protect donor privacy and minimize cross-service identity correlation. proposal
  9. CL-39-04.09 Measure error, reversal, demographic impact, and political selectivity. proposal
4 evidence units
Report-derived argument

The protected domain begins before a person speaks

CL-40-01
As of
Next review
Claim kind
global synthesis

Section summary: Freedom of expression protects an outward act, but speech begins in a less visible process: attention, reading, memory, uncertainty, imagination, comparison, and belief formation. International human-rights law recognizes an internal domain of thought and conscience that receives especially strong protection. Modern technology makes that domain easier to

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 3 inventoried assertions
  1. CL-40-01.01 Freedom of expression protects an outward act, but speech begins in a less visible process: attention, reading, memory, uncertainty, imagination, comparison, and belief formation. International human-rights law recognizes an internal domain of thought and conscience that receives especially strong protection. Modern technology makes that domain easier to observe indirectly through search, conversation, behavior, and neural signals. descriptive assertion
  2. CL-40-01.02 Cognitive liberty names the combined interest in preserving this process. It includes the right not to have intimate inferences extracted or used without justification, the right to investigate lawful ideas in private, the right not to be covertly manipulated, and the right to retain continuity of identity and memory. It also includes access to the informational conditions needed to make a choice rather than merely the absence of direct force. descriptive assertion
  3. CL-40-01.03 Foundational claim: a society cannot reliably protect free expression after allowing institutions to monitor, narrow, or manipulate the private process from which expression emerges. descriptive assertion
Report-derived argument

A charter needs rights, duties, and limits

CL-40-02
As of
Next review
Claim kind
global synthesis

Section summary: A purely negative rule—do not read minds—is too narrow. Most present threats arise from ordinary data systems: query logs, recommendation profiles, workplace monitoring, identity graphs, advertising inferences, AI conversation histories, and mandatory educational platforms. A complete framework therefore combines privacy, access, agency, and due process.

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-OHCHR-ICCPR International Covenant on Civil and Political Rights — Article 18 · Primary law or court record
  • SRC-OHCHR-A76-380 UN Special Rapporteur report A/76/380 — Freedom of thought · Official oversight record
Inspect 9 inventoried assertions
  1. CL-40-02.01 A purely negative rule—do not read minds—is too narrow. Most present threats arise from ordinary data systems: query logs, recommendation profiles, workplace monitoring, identity graphs, advertising inferences, AI conversation histories, and mandatory educational platforms. A complete framework therefore combines privacy, access, agency, and due process. descriptive assertion
  2. CL-40-02.02 The right is also not unlimited immunity from conduct law. True threats, fraud, exploitation, unauthorized access, and direct facilitation of serious harm can be restricted. The safeguard is to regulate the harmful act or material capability without converting broad populations’ reading, thought, or curiosity into a permanent object of surveillance. descriptive assertion
  3. CL-40-02.03 Protection Institutional duty descriptive assertion
  4. CL-40-02.04 Mental privacy Minimize neural, behavioral, and inferential data; require strong authority for access. descriptive assertion
  5. CL-40-02.05 Freedom of inquiry Do not treat lawful questions as proof of belief or intent. descriptive assertion
  6. CL-40-02.06 Algorithmic autonomy Disclose and control profiling, personalization, and manipulation. descriptive assertion
  7. CL-40-02.07 Plural information Preserve access, interoperability, translation, and independent sources. descriptive assertion
  8. CL-40-02.08 Cognitive tools Protect voluntary use and refusal, especially in power-imbalanced settings. descriptive assertion
  9. CL-40-02.09 Due process Give notice, explanation, correction, appeal, and remedy. descriptive assertion
Scenario analysis

Worldwide protection must avoid a global orthodoxy

CL-40-03
As of
Next review
Claim kind
global synthesis

Section summary: A global rights floor should not become one institution’s authority to define acceptable thought for every culture and language. Human dignity, mental privacy, lawful inquiry, and freedom from coercion can be protected while allowing plural moral, religious, political, and scientific traditions to disagree. The architecture should decentralize interpretive

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-OHCHR-A76-380 UN Special Rapporteur report A/76/380 — Freedom of thought · Official oversight record
  • SRC-UNESCO-AI-ETHICS UNESCO Recommendation on the Ethics of Artificial Intelligence · Primary law or court record
Inspect 6 inventoried assertions
  1. CL-40-03.01 A global rights floor should not become one institution’s authority to define acceptable thought for every culture and language. Human dignity, mental privacy, lawful inquiry, and freedom from coercion can be protected while allowing plural moral, religious, political, and scientific traditions to disagree. scenario
  2. CL-40-03.02 The architecture should decentralize interpretive power. International instruments can state minimum protections; national and local institutions can implement them; independent courts and civil society can contest overreach; open technical systems can offer alternatives; and communities can preserve their own languages and records. The test is whether a person retains meaningful capacity to seek, compare, dissent, and change their mind. scenario
  3. CL-40-03.03 Global floor Thought, privacy, inquiry, integrity, and remedy. scenario
  4. CL-40-03.04 Local plurality Different cultures may reason and organize differently. scenario
  5. CL-40-03.05 Independent friction Courts, media, libraries, auditors, and civil society. scenario
  6. CL-40-03.06 Technical exit Open tools, portability, local computation, and interoperable networks. scenario
Site policy proposal

Implementation agenda for the next decade

CL-40-04
As of
Next review
Claim kind
policy proposal

Section summary: Recognize neural data, intimate behavioral inferences, search histories, and AI conversations as highly sensitive cognitive records. Protect anonymous and pseudonymous lawful inquiry and prohibit universal identity linkage as a default condition of access. Ban non-consensual cognitive manipulation, punitive emotion inference, and compelled neurotechnology

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-OHCHR-A76-380 UN Special Rapporteur report A/76/380 — Freedom of thought · Official oversight record
  • SRC-UNESCO-AI-ETHICS UNESCO Recommendation on the Ethics of Artificial Intelligence · Primary law or court record
  • SRC-UAIX-COGNITIVE-LIBERTY-CHARTER UAIX — Cognitive Liberty Charter Draft · Professional or project methodology
  • SRC-OHCHR-ICCPR International Covenant on Civil and Political Rights — Article 18 · Primary law or court record
Inspect 9 inventoried assertions
  1. CL-40-04.01 The framework is successful only when ordinary people can exercise it. Rights should therefore be translated into product requirements, procurement rules, professional ethics, technical standards, court remedies, and measurable public evidence. proposal
  2. CL-40-04.02 Recognize neural data, intimate behavioral inferences, search histories, and AI conversations as highly sensitive cognitive records. proposal
  3. CL-40-04.03 Protect anonymous and pseudonymous lawful inquiry and prohibit universal identity linkage as a default condition of access. proposal
  4. CL-40-04.04 Ban non-consensual cognitive manipulation, punitive emotion inference, and compelled neurotechnology in workplaces, schools, policing, and detention. proposal
  5. CL-40-04.05 Require least-intrusive safety design, defined harms, measurable effectiveness, purpose limits, deletion, independent review, and sunset clauses. proposal
  6. CL-40-04.06 Protect encryption, local AI, private computation, open standards, portability, and plural model ecosystems. proposal
  7. CL-40-04.07 Guarantee discoverability due process when automated systems materially suppress lawful information or opportunity. proposal
  8. CL-40-04.08 Fund libraries, universities, journalists, translators, archives, diaspora networks, and public-interest technology as cognitive-liberty infrastructure. proposal
  9. CL-40-04.09 Create independent audits and remedies that test system behavior rather than accepting policy promises alone. proposal
4 evidence units
Evidence + editorial analysis

Censorship targets the institution around the reporter

CL-41-01
As of
Next review
Claim kind
global synthesis

Section summary: A state does not need to pre-approve every sentence if it can control the conditions of reporting. Licensing can make publication revocable. Friendly owners can replace editors. Defamation, security, extremism, or foreign-agent laws can impose legal risk. Tax audits and advertising pressure can drain resources. Police, militias, or coordinated harassment

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 3 inventoried assertions
  1. CL-41-01.01 A state does not need to pre-approve every sentence if it can control the conditions of reporting. Licensing can make publication revocable. Friendly owners can replace editors. Defamation, security, extremism, or foreign-agent laws can impose legal risk. Tax audits and advertising pressure can drain resources. Police, militias, or coordinated harassment can make ordinary reporting physically dangerous. inference or risk
  2. CL-41-01.02 These mechanisms reinforce one another. A newsroom weakened financially becomes more dependent on one funder or platform. A journalist who expects device seizure cannot promise source confidentiality. A broadcaster denied spectrum or app-store access may technically remain legal but lose its audience. The system’s effectiveness lies in distributing responsibility so no single act appears to be the whole censorship programme. inference or risk
  3. CL-41-01.03 Institutional test: press freedom should be measured across ownership, licensing, law, finance, distribution, safety, source protection, and access to records—not only by whether a formal censorship office exists. descriptive assertion
Evidence + editorial analysis

Exile preserves reporting but creates new dependencies

CL-41-02
As of
Next review
Claim kind
global synthesis

Section summary: When domestic reporting becomes impossible, journalists relocate and continue through foreign publishers, diaspora networks, secure messaging, mirrored sites, video platforms, newsletters, and collaboration with reporters who remain inside. Exile can protect editorial freedom and preserve records that would otherwise disappear. It also creates distance from

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-41-02.01 When domestic reporting becomes impossible, journalists relocate and continue through foreign publishers, diaspora networks, secure messaging, mirrored sites, video platforms, newsletters, and collaboration with reporters who remain inside. Exile can protect editorial freedom and preserve records that would otherwise disappear. descriptive assertion
  2. CL-41-02.02 It also creates distance from sources, unstable immigration status, donor dependence, language barriers, payment problems, platform chokepoints, and host-country security concerns. Origin states may target relatives, deploy spyware, flood audiences with impersonations, or pressure host governments. A resilient exile outlet therefore needs local reporting networks, transparent correction, diversified distribution, secure source handling, and a funding model that does not force editorial conformity. inference or risk
  3. CL-41-02.03 Risk Protective practice descriptive assertion
  4. CL-41-02.04 Loss of domestic access Trusted inside networks and distributed verification descriptive assertion
  5. CL-41-02.05 Platform blocking Mirrors, newsletters, multiple protocols, and offline channels descriptive assertion
  6. CL-41-02.06 Transnational repression Threat assessment, host-state protection, and family safeguards descriptive assertion
  7. CL-41-02.07 Funding capture Diverse donors, reserves, and editorial firewalls descriptive assertion
  8. CL-41-02.08 Credibility attacks Visible sources, corrections, methods, and uncertainty descriptive assertion
Evidence + editorial analysis

Verification becomes more important when access is restricted

CL-41-03
As of
Next review
Claim kind
global synthesis

Section summary: Censorship creates an evidentiary paradox. Independent reporters have less direct access precisely when propaganda and fabricated media increase. User-generated video, satellite imagery, leaked records, witness testimony, and open-source intelligence can fill gaps, but each source has limitations and may expose contributors. Verification should separate

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 6 inventoried assertions
  1. CL-41-03.01 Censorship creates an evidentiary paradox. Independent reporters have less direct access precisely when propaganda and fabricated media increase. User-generated video, satellite imagery, leaked records, witness testimony, and open-source intelligence can fill gaps, but each source has limitations and may expose contributors. inference or risk
  2. CL-41-03.02 Verification should separate acquisition from publication. Newsrooms can preserve original bytes, metadata, source communications, and chain-of-custody records in a protected system while publishing redacted evidence and an explanation of confidence. Synthetic-media detection is only one signal; contextual corroboration, chronology, geolocation, source history, and independent witnesses remain necessary. descriptive assertion
  3. CL-41-03.03 Receive Use secure, compartmentalized source channels. descriptive assertion
  4. CL-41-03.04 Preserve Keep original files, metadata, and custody records. descriptive assertion
  5. CL-41-03.05 Verify Corroborate place, time, identity, and sequence. descriptive assertion
  6. CL-41-03.06 Publish Explain confidence, redaction, and unresolved uncertainty. descriptive assertion
Site policy proposal

A worldwide support system for independent journalism

CL-41-04
As of
Next review
Claim kind
policy proposal

Section summary: Protect confidential sources, encrypted communication, anonymous publication, and secure whistleblowing channels. Provide rapid legal, physical-security, relocation, trauma, and digital-forensics support. Fund core operations and emergency reserves without demanding editorial alignment. Require platforms and infrastructure providers to disclose government

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-41-04.01 Independent journalism is not merely content. It is a network of people, records, tools, institutions, and trust. Protecting the article while leaving that network exposed is not sufficient. proposal
  2. CL-41-04.02 Protect confidential sources, encrypted communication, anonymous publication, and secure whistleblowing channels. proposal
  3. CL-41-04.03 Provide rapid legal, physical-security, relocation, trauma, and digital-forensics support. proposal
  4. CL-41-04.04 Fund core operations and emergency reserves without demanding editorial alignment. proposal
  5. CL-41-04.05 Require platforms and infrastructure providers to disclose government demands and give notice and appeal where safe. proposal
  6. CL-41-04.06 Build cross-border archives so investigations and evidence survive the closure of one outlet. proposal
  7. CL-41-04.07 Support translation, audience research, and low-bandwidth distribution for people inside restricted networks. proposal
  8. CL-41-04.08 Investigate spyware, harassment, financial exclusion, and threats against relatives as parts of the censorship system. proposal
  9. CL-41-04.09 Maintain rigorous correction and provenance practices so resistance to censorship does not become exemption from evidence. proposal
4 evidence units
Evidence + editorial analysis

Censorship can operate by redefining reality

CL-42-01
As of
Next review
Claim kind
global synthesis

Section summary: Authorities often govern language before governing conclusions. A war can be renamed an operation, a prison a reeducation facility, a civilian death collateral damage, or a political opponent a threat to social harmony. The new term may contain some technical truth, but it can also remove agency, suffering, and responsibility from public understanding.

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
  • SRC-OHCHR-RABAT-PLAN OHCHR — Rabat Plan of Action · Professional or project methodology
Inspect 3 inventoried assertions
  1. CL-42-01.01 Authorities often govern language before governing conclusions. A war can be renamed an operation, a prison a reeducation facility, a civilian death collateral damage, or a political opponent a threat to social harmony. The new term may contain some technical truth, but it can also remove agency, suffering, and responsibility from public understanding. inference or risk
  2. CL-42-01.02 Controlled vocabulary works through repetition and consequence. Schools, media, public agencies, and platforms normalize the approved phrasing. Direct terms become professionally risky, algorithmically downranked, or legally punishable. Eventually people may use the euphemism even in private because it is the only language that circulates widely. inference or risk
  3. CL-42-01.03 Semantic warning sign: watch for language rules that make criticism less precise while making institutional power harder to name. inference or risk
Evidence + editorial analysis

Not every change in terminology is coercive

CL-42-02
As of
Next review
Claim kind
global synthesis

Section summary: Communities revise language to become more accurate, reduce needless stigma, reflect scientific knowledge, or respect how people describe themselves. Technical fields also need controlled vocabularies so records can be compared. Cognitive liberty does not require freezing language or treating every social norm as censorship. The distinction is process and

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-OHCHR-RABAT-PLAN OHCHR — Rabat Plan of Action · Professional or project methodology
  • SRC-CLEAR-GLOBAL CLEAR Global — language access and crisis communication · Professional or project methodology
Inspect 7 inventoried assertions
  1. CL-42-02.01 Communities revise language to become more accurate, reduce needless stigma, reflect scientific knowledge, or respect how people describe themselves. Technical fields also need controlled vocabularies so records can be compared. Cognitive liberty does not require freezing language or treating every social norm as censorship. descriptive assertion
  2. CL-42-02.02 The distinction is process and power. Is the term open to debate? Can speakers explain why they reject it? Does the change improve descriptive accuracy? Are penalties proportionate? Is one institution protecting vulnerable people or protecting itself from scrutiny? A language norm becomes coercive when dissent from the vocabulary is punished regardless of context and no accountable route exists to challenge the definition. descriptive assertion
  3. CL-42-02.03 Question Open linguistic norm Control mechanism descriptive assertion
  4. CL-42-02.04 Purpose Precision, inclusion, or shared meaning Concealment or ideological obedience descriptive assertion
  5. CL-42-02.05 Authority Plural, revisable, and explained Central, secret, or punitive descriptive assertion
  6. CL-42-02.06 Context Meaning depends on use Word alone triggers action descriptive assertion
  7. CL-42-02.07 Remedy Debate and correction remain possible Challenge creates further penalty descriptive assertion
Evidence + editorial analysis

Coded language preserves speech but transfers the burden to users

CL-42-03
As of
Next review
Claim kind
global synthesis

Section summary: People respond to forbidden words with allegory, jokes, abbreviations, misspellings, homophones, images, historical analogy, and “algospeak.” These methods create a temporary commons among those who understand the code and can frustrate automated filters. The adaptation has costs. Meaning becomes harder for outsiders, disabled users, translators, archives,

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-CLEAR-GLOBAL CLEAR Global — language access and crisis communication · Professional or project methodology
  • SRC-UNESCO-AI-ETHICS UNESCO Recommendation on the Ethics of Artificial Intelligence · Primary law or court record
Inspect 6 inventoried assertions
  1. CL-42-03.01 People respond to forbidden words with allegory, jokes, abbreviations, misspellings, homophones, images, historical analogy, and “algospeak.” These methods create a temporary commons among those who understand the code and can frustrate automated filters. descriptive assertion
  2. CL-42-03.02 The adaptation has costs. Meaning becomes harder for outsiders, disabled users, translators, archives, and future researchers to recover. Constant code changes fracture communities and make misinformation easier to spread. The burden of evasion falls on the censored speaker while the institution avoids explaining its rule. Resistance language is evidence of ingenuity, but it is not a substitute for rights-respecting governance. descriptive assertion
  3. CL-42-03.03 Ban A term becomes risky or invisible. inference or risk
  4. CL-42-03.04 Adapt Users invent metaphor, spelling, or symbols. descriptive assertion
  5. CL-42-03.05 Detect Classifiers learn the new code. descriptive assertion
  6. CL-42-03.06 Drift Meaning fragments and the cycle repeats. descriptive assertion
Site policy proposal

Protect linguistic freedom in AI and public institutions

CL-42-04
As of
Next review
Claim kind
policy proposal

Section summary: Define prohibited conduct and context rather than maintaining secret lists of politically dangerous words. Publish the language categories and examples used by automated moderation systems. Test classifiers across dialects, minority languages, reclaimed terms, quotation, satire, journalism, and historical discussion. Preserve original terminology alongside

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-CLEAR-GLOBAL CLEAR Global — language access and crisis communication · Professional or project methodology
  • SRC-UNESCO-AI-ETHICS UNESCO Recommendation on the Ethics of Artificial Intelligence · Primary law or court record
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
  • SRC-OHCHR-RABAT-PLAN OHCHR — Rabat Plan of Action · Professional or project methodology
Inspect 9 inventoried assertions
  1. CL-42-04.01 Language is not merely a delivery mechanism for thought. It supplies the distinctions through which people recognize power, harm, identity, and possibility. Protecting cognitive liberty therefore includes protecting the ability to name reality accurately. proposal
  2. CL-42-04.02 Define prohibited conduct and context rather than maintaining secret lists of politically dangerous words. proposal
  3. CL-42-04.03 Publish the language categories and examples used by automated moderation systems. proposal
  4. CL-42-04.04 Test classifiers across dialects, minority languages, reclaimed terms, quotation, satire, journalism, and historical discussion. proposal
  5. CL-42-04.05 Preserve original terminology alongside translations, summaries, and later editorial labels. proposal
  6. CL-42-04.06 Let people inspect, correct, and appeal language-based risk or content classifications. proposal
  7. CL-42-04.07 Protect scholars, journalists, artists, translators, and communities who document changing political vocabulary. proposal
  8. CL-42-04.08 Require generative AI to disclose uncertainty and preserve alternative terms rather than silently normalizing one official lexicon. proposal
  9. CL-42-04.09 Archive euphemisms and coded language as evidence of the conditions under which people communicated. proposal
4 evidence units
Evidence + editorial analysis

Humor converts private recognition into common knowledge

CL-43-01
As of
Next review
Claim kind
global synthesis

Section summary: A political joke does more than communicate a proposition. It tests whether an audience recognizes the gap between official language and lived experience. Laughter signals that the perception is shared. Under conditions of preference falsification, this can be politically important because people discover that their private doubts are not isolated. Satire

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-FIRST-MONDAY-POLITICAL-HUMOR First Monday — political satire and coded online expression · Original peer-reviewed research
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 3 inventoried assertions
  1. CL-43-01.01 A political joke does more than communicate a proposition. It tests whether an audience recognizes the gap between official language and lived experience. Laughter signals that the perception is shared. Under conditions of preference falsification, this can be politically important because people discover that their private doubts are not isolated. descriptive assertion
  2. CL-43-01.02 Satire also reduces the dignity on which authoritarian power often depends. A slogan demands repetition; a joke rearranges it. A monumental image becomes a meme. A claim of inevitability becomes absurd. This is why punishment may focus less on factual falsity than on disrespect, insult, humiliation, or the loss of narrative control. inference or risk
  3. CL-43-01.03 Civic function: humor can make hidden disagreement socially visible before people are willing to state the disagreement directly. descriptive assertion
Evidence + editorial analysis

Satire is powerful because meaning depends on context

CL-43-02
As of
Next review
Claim kind
global synthesis

Section summary: The same ambiguity that protects humor can cause harm. A racist statement can be defended as a joke; a threat can be disguised as irony; a parody can be mistaken for authentic speech. A rights-respecting system must examine speaker, audience, target, intent, content, reach, likelihood of harm, and surrounding context rather than treating “satire” as an

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
  • SRC-OHCHR-RABAT-PLAN OHCHR — Rabat Plan of Action · Professional or project methodology
Inspect 7 inventoried assertions
  1. CL-43-02.01 The same ambiguity that protects humor can cause harm. A racist statement can be defended as a joke; a threat can be disguised as irony; a parody can be mistaken for authentic speech. A rights-respecting system must examine speaker, audience, target, intent, content, reach, likelihood of harm, and surrounding context rather than treating “satire” as an automatic exemption. descriptive assertion
  2. CL-43-02.02 Authorities abuse this uncertainty when they collapse ridicule of officials into threats to the state or criminalize “insult” without a concrete harm standard. Platforms can reproduce the problem when automated systems classify the literal words or image while missing inversion, quotation, parody, reclaimed language, and counterspeech. descriptive assertion
  3. CL-43-02.03 Question Why it matters descriptive assertion
  4. CL-43-02.04 What is being imitated? Parody often repeats the target’s language to criticize it. descriptive assertion
  5. CL-43-02.05 Who is the target? Mocking power differs from coordinated harassment of a vulnerable person. descriptive assertion
  6. CL-43-02.06 What action is urged? Ridicule differs from a credible threat or incitement. descriptive assertion
  7. CL-43-02.07 How is it framed? Caption, account history, and audience cues may reverse literal meaning. inference or risk
Evidence + editorial analysis

The algorithmic censor is unusually bad at irony

CL-43-03
As of
Next review
Claim kind
global synthesis

Section summary: Automated moderation is built for scale and repeatability. Satire is built from violation of expectation. Models trained on decontextualized examples may over-remove political cartoons, war jokes, minority vernacular, or quoted extremist slogans. Other systems may under-remove harassment because the abusive meaning is encoded through a community-specific

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-OHCHR-RABAT-PLAN OHCHR — Rabat Plan of Action · Professional or project methodology
  • SRC-SANTA-CLARA-PRINCIPLES Santa Clara Principles on Transparency and Accountability in Content Moderation · Professional or project methodology
Inspect 6 inventoried assertions
  1. CL-43-03.01 Automated moderation is built for scale and repeatability. Satire is built from violation of expectation. Models trained on decontextualized examples may over-remove political cartoons, war jokes, minority vernacular, or quoted extremist slogans. Other systems may under-remove harassment because the abusive meaning is encoded through a community-specific joke. inference or risk
  2. CL-43-03.02 Generative AI adds another risk: it can explain a sanitized political history while refusing to reproduce the humor through which people contested it. Translation systems may flatten wordplay or choose an official interpretation. Search and recommendation systems may demote satire as low quality, unsafe, or advertiser-unfriendly, reducing its public reach without a formal ban. inference or risk
  3. CL-43-03.03 Literal text The classifier sees prohibited words. descriptive assertion
  4. CL-43-03.04 Pragmatic meaning The audience recognizes parody or inversion. descriptive assertion
  5. CL-43-03.05 Policy action Automated enforcement ignores the gap. descriptive assertion
  6. CL-43-03.06 Chill Creators simplify, code, or stop speaking. descriptive assertion
Site policy proposal

Protect satire without immunizing abuse

CL-43-04
As of
Next review
Claim kind
policy proposal

Section summary: Reject laws that criminalize ridicule of officials, institutions, flags, or state narratives without a concrete and proportionate harm basis. Use context, intent, target, reach, and likelihood tests for incitement and threats. Build parody, quotation, counterspeech, and artistic-context review into moderation appeals. Test models and classifiers with

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-OHCHR-RABAT-PLAN OHCHR — Rabat Plan of Action · Professional or project methodology
  • SRC-SANTA-CLARA-PRINCIPLES Santa Clara Principles on Transparency and Accountability in Content Moderation · Professional or project methodology
  • SRC-FIRST-MONDAY-POLITICAL-HUMOR First Monday — political satire and coded online expression · Original peer-reviewed research
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 8 inventoried assertions
  1. CL-43-04.01 A free society does not require everyone to find a joke funny. It requires that offense, embarrassment, or loss of prestige not automatically become a state or platform justification for silencing political imagination. proposal
  2. CL-43-04.02 Reject laws that criminalize ridicule of officials, institutions, flags, or state narratives without a concrete and proportionate harm basis. proposal
  3. CL-43-04.03 Use context, intent, target, reach, and likelihood tests for incitement and threats. proposal
  4. CL-43-04.04 Build parody, quotation, counterspeech, and artistic-context review into moderation appeals. proposal
  5. CL-43-04.05 Test models and classifiers with multilingual jokes, memes, dialects, reclaimed terms, and historical examples. proposal
  6. CL-43-04.06 Give creators reasons and restoration remedies when automated enforcement misreads context. proposal
  7. CL-43-04.07 Preserve censored humor as cultural and historical evidence, including the context needed to understand it. proposal
  8. CL-43-04.08 Protect comedians, cartoonists, meme creators, and satirists from violence and transnational retaliation. proposal
4 evidence units
Evidence + editorial analysis

Reading is part of thought formation, not merely content consumption

CL-44-01
As of
Next review
Claim kind
global synthesis

Section summary: A public statement is usually edited for an audience. A reading choice may be uncertain, private, contradictory, or experimental. People read arguments they reject, investigate stigmatized experiences, compare political systems, study crimes, and explore identity before they know what they believe. Treating the record as a stable profile mistakes inquiry

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 3 inventoried assertions
  1. CL-44-01.01 A public statement is usually edited for an audience. A reading choice may be uncertain, private, contradictory, or experimental. People read arguments they reject, investigate stigmatized experiences, compare political systems, study crimes, and explore identity before they know what they believe. Treating the record as a stable profile mistakes inquiry for conclusion. inference or risk
  2. CL-44-01.02 Intellectual privacy protects this unfinished space. It enables dissent, scholarship, creativity, faith, health research, and self-understanding by reducing the anticipated cost of asking. When people believe a library, employer, government, family member, or platform will inspect their reading, they may choose safer material and lose access to the evidence needed to form an independent view. inference or risk
  3. CL-44-01.03 Presumption: lawful reading is not evidence of endorsement, dangerousness, disloyalty, illness, or character without additional, context-specific proof. inference or risk
Evidence + editorial analysis

Digital reading creates a much richer surveillance record

CL-44-02
As of
Next review
Claim kind
global synthesis

Section summary: A physical library may know that a book was borrowed and can often delete the record after return. A digital service can record the query, result list, selected edition, passage viewed, reading speed, highlight, note, device, location, referral, and later recommendation response. Multiple services can combine these signals into an inferred psychological or

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-ALA-LMS-PRIVACY American Library Association — privacy guidelines for library management systems · Professional or project methodology
  • SRC-SCOTUS-PICO-1982 Board of Education v. Pico — bound U.S. Reports · Primary law or court record
Inspect 8 inventoried assertions
  1. CL-44-02.01 A physical library may know that a book was borrowed and can often delete the record after return. A digital service can record the query, result list, selected edition, passage viewed, reading speed, highlight, note, device, location, referral, and later recommendation response. Multiple services can combine these signals into an inferred psychological or political profile. inference or risk
  2. CL-44-02.02 Personalization and synchronization may benefit readers, but the same data can be breached, sold, compelled, used for advertising, or applied to employment, insurance, credit, policing, and content restriction. Privacy cannot depend only on a promise not to misuse an indefinitely retained record. Architecture should reduce the amount that exists. inference or risk
  3. CL-44-02.03 Data Default protection descriptive assertion
  4. CL-44-02.04 Search and browse history Ephemeral or short retention; not linked across services descriptive assertion
  5. CL-44-02.05 Borrowing or purchase record Retain only for the transaction and user-selected history descriptive assertion
  6. CL-44-02.06 Highlights and notes End-to-end protected or local where practical descriptive assertion
  7. CL-44-02.07 Interest inference No sensitive-trait profile without explicit purpose and consent descriptive assertion
  8. CL-44-02.08 Government access Specific legal process with heightened protection for expressive records descriptive assertion
Evidence + editorial analysis

Privacy should follow the function across media

CL-44-03
As of
Next review
Claim kind
global synthesis

Section summary: Library confidentiality loses much of its value if the same inquiry becomes exposed when conducted through an e-book platform, search engine, university database, video archive, or AI assistant. The protected interest is not the building or format; it is the person’s ability to receive and develop ideas without creating an unnecessary surveillance record.

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-SCOTUS-PICO-1982 Board of Education v. Pico — bound U.S. Reports · Primary law or court record
  • SRC-OHCHR-ICCPR International Covenant on Civil and Political Rights — Article 18 · Primary law or court record
Inspect 6 inventoried assertions
  1. CL-44-03.01 Library confidentiality loses much of its value if the same inquiry becomes exposed when conducted through an e-book platform, search engine, university database, video archive, or AI assistant. The protected interest is not the building or format; it is the person’s ability to receive and develop ideas without creating an unnecessary surveillance record. descriptive assertion
  2. CL-44-03.02 Different systems need different operations. A retailer may need payment records. A university may need licensed access. A service may need abuse prevention. These functions do not require every page view or question to become an indefinite, identity-linked profile. Privacy partitioning, local history, anonymous credentials, aggregate measurement, and strict purpose separation can preserve service functionality. inference or risk
  3. CL-44-03.03 Ask Keep the query separate from legal identity. descriptive assertion
  4. CL-44-03.04 Receive Deliver the work without cross-context profiling. descriptive assertion
  5. CL-44-03.05 Remember Store history only when the reader chooses. descriptive assertion
  6. CL-44-03.06 Disclose Require narrow process and notify where lawful. descriptive assertion
Site policy proposal

A worldwide inquiry-privacy standard

CL-44-04
As of
Next review
Claim kind
policy proposal

Section summary: Protect reading, search, viewing, borrowing, annotation, and AI-query records as highly sensitive intellectual data. Collect only what the immediate transaction requires and delete it on a fixed, short schedule. Make saved history, personalization, and cross-device synchronization opt-in and easy to erase. Prohibit use of lawful reading to infer

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-SCOTUS-PICO-1982 Board of Education v. Pico — bound U.S. Reports · Primary law or court record
  • SRC-OHCHR-ICCPR International Covenant on Civil and Political Rights — Article 18 · Primary law or court record
  • SRC-RICHARDS-INTELLECTUAL-PRIVACY Neil Richards — Intellectual Privacy · Original preprint or working paper
  • SRC-ALA-LMS-PRIVACY American Library Association — privacy guidelines for library management systems · Professional or project methodology
Inspect 9 inventoried assertions
  1. CL-44-04.01 The right to read privately is not a demand that all transactions be anonymous or that unlawful conduct be immune from investigation. It is a demand that the ordinary process of becoming informed not become a permanent dossier by default. proposal
  2. CL-44-04.02 Protect reading, search, viewing, borrowing, annotation, and AI-query records as highly sensitive intellectual data. proposal
  3. CL-44-04.03 Collect only what the immediate transaction requires and delete it on a fixed, short schedule. proposal
  4. CL-44-04.04 Make saved history, personalization, and cross-device synchronization opt-in and easy to erase. proposal
  5. CL-44-04.05 Prohibit use of lawful reading to infer eligibility, employability, insurability, creditworthiness, politics, religion, or dangerousness without a separate lawful basis and meaningful due process. proposal
  6. CL-44-04.06 Require particularized legal authority for government access and challenge reverse or keyword dragnets. proposal
  7. CL-44-04.07 Bind vendors serving schools, libraries, universities, and newsrooms to the same confidentiality duties as the institution. proposal
  8. CL-44-04.08 Provide anonymous or pseudonymous access paths for public-interest information. proposal
  9. CL-44-04.09 Publish transparency reports and notify readers of demands whenever law permits. proposal
4 evidence units
Evidence + editorial analysis

Private speech is where public thought is prepared

CL-45-01
As of
Next review
Claim kind
global synthesis

Section summary: People rarely arrive at complex beliefs alone and fully formed. They test language with friends, consult colleagues, admit uncertainty, ask embarrassing questions, and revise their views in small trusted groups. Journalists develop sources, workers discuss conditions, families interpret events, and communities decide whether a public concern is shared. When

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 3 inventoried assertions
  1. CL-45-01.01 People rarely arrive at complex beliefs alone and fully formed. They test language with friends, consult colleagues, admit uncertainty, ask embarrassing questions, and revise their views in small trusted groups. Journalists develop sources, workers discuss conditions, families interpret events, and communities decide whether a public concern is shared. descriptive assertion
  2. CL-45-01.02 When every conversation may be retained, searched, forwarded, breached, or scored, this preparatory space changes. People avoid names, simplify disagreements, stop contacting risky associates, or remain silent. Surveillance therefore affects expression upstream: it can prevent the relationships and shared understanding from which public speech and collective action emerge. inference or risk
  3. CL-45-01.03 Social principle: privacy is not isolation. It is the condition that lets people form trusted relationships and produce honest common knowledge. descriptive assertion
Evidence + editorial analysis

Metadata can reveal the structure of thought communities

CL-45-02
As of
Next review
Claim kind
global synthesis

Section summary: Encryption can conceal message content while leaving metadata: who communicated, when, from which devices, for how long, and in what group. At scale, these records reveal networks of journalists, activists, patients, religious communities, lawyers, researchers, and political organizations. AI can classify roles, central figures, and likely topics by

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-45-02.01 Encryption can conceal message content while leaving metadata: who communicated, when, from which devices, for how long, and in what group. At scale, these records reveal networks of journalists, activists, patients, religious communities, lawyers, researchers, and political organizations. AI can classify roles, central figures, and likely topics by combining metadata with location and public information. descriptive assertion
  2. CL-45-02.02 Metadata is also necessary for routing, abuse prevention, billing, and reliability. The design question is whether the service stores more than the immediate function requires, retains it indefinitely, links it to other accounts, or provides it for unrelated profiling. Partitioned routing, sealed-sender designs, private contact discovery, short retention, and aggregate telemetry can reduce exposure. descriptive assertion
  3. CL-45-02.03 Layer What can be exposed Protection descriptive assertion
  4. CL-45-02.04 Content Words, files, and media End-to-end encryption descriptive assertion
  5. CL-45-02.05 Metadata Participants, timing, groups, location Minimization and private routing descriptive assertion
  6. CL-45-02.06 Endpoint Notifications, screenshots, local storage Device security and user controls descriptive assertion
  7. CL-45-02.07 Backup Plaintext copies and keys Encrypted, optional, recoverable design descriptive assertion
  8. CL-45-02.08 Inference Relationships, roles, interests, risk scores Purpose limits and bans on sensitive profiling inference or risk
Evidence + editorial analysis

Targeted investigation is different from universal observability

CL-45-03
As of
Next review
Claim kind
global synthesis

Section summary: Governments have legitimate authority to investigate serious crime, threats, exploitation, and espionage under law. That need does not establish that every message should be scannable or every conversation graph indefinitely available. A universal access mechanism creates a population-wide vulnerability whose target list can change without rebuilding the

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 6 inventoried assertions
  1. CL-45-03.01 Governments have legitimate authority to investigate serious crime, threats, exploitation, and espionage under law. That need does not establish that every message should be scannable or every conversation graph indefinitely available. A universal access mechanism creates a population-wide vulnerability whose target list can change without rebuilding the system. descriptive assertion
  2. CL-45-03.02 Client-side scanning is often presented as preserving encryption because inspection occurs before encryption. From the user’s perspective, however, private content is still searched by a rule controlled elsewhere. Less intrusive alternatives include targeted device seizure with judicial authority, undercover investigation, victim reports, account-level evidence, and privacy-preserving safety checks that do not report ordinary lawful content. descriptive assertion
  3. CL-45-03.03 Specific harm Identify the event and evidence threshold. descriptive assertion
  4. CL-45-03.04 Targeted authority Use individualized, reviewable legal process. descriptive assertion
  5. CL-45-03.05 Minimized collection Avoid bulk content and contact-graph retention. descriptive assertion
  6. CL-45-03.06 Remedy Audit access, notify later where possible, and correct errors. descriptive assertion
Site policy proposal

Defend the private sphere by design and law

CL-45-04
As of
Next review
Claim kind
policy proposal

Section summary: Protect end-to-end encryption and reject universal backdoors or remotely expandable client-side scanning mandates. Minimize metadata, logs, contact discovery, location, and cross-service identifiers. Make encrypted backups and disappearing-message controls understandable and user-directed. Require particularized legal process, access logging, minimization,

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-45-04.01 The goal is not a world in which harmful conduct becomes uninvestigable. It is a world in which ordinary human trust does not require accepting a permanent observer in every conversation. proposal
  2. CL-45-04.02 Protect end-to-end encryption and reject universal backdoors or remotely expandable client-side scanning mandates. proposal
  3. CL-45-04.03 Minimize metadata, logs, contact discovery, location, and cross-service identifiers. proposal
  4. CL-45-04.04 Make encrypted backups and disappearing-message controls understandable and user-directed. proposal
  5. CL-45-04.05 Require particularized legal process, access logging, minimization, and independent review for compelled data. proposal
  6. CL-45-04.06 Protect journalist-source, lawyer-client, medical, religious, and research confidentiality. proposal
  7. CL-45-04.07 Prohibit sensitive political, religious, medical, and psychological profiling from private conversation graphs. proposal
  8. CL-45-04.08 Support open protocol review, reproducible builds, secure updates, and independent security research. proposal
  9. CL-45-04.09 Teach realistic threat models without implying that any tool provides perfect anonymity. proposal
4 evidence units
Evidence + editorial analysis

Translation is selection before it is wording

CL-46-01
As of
Next review
Claim kind
global synthesis

Section summary: The first censorship decision may be whether a work is translated at all. Institutions can translate foreign praise while ignoring criticism, select sensational fragments that confirm a domestic narrative, or publish only commentary about a source that readers cannot inspect. This creates an information barrier without formally banning the original.

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-CLEAR-GLOBAL CLEAR Global — language access and crisis communication · Professional or project methodology
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 3 inventoried assertions
  1. CL-46-01.01 The first censorship decision may be whether a work is translated at all. Institutions can translate foreign praise while ignoring criticism, select sensational fragments that confirm a domestic narrative, or publish only commentary about a source that readers cannot inspect. This creates an information barrier without formally banning the original. inference or risk
  2. CL-46-01.02 Selection also shapes urgency and legitimacy. Which witnesses receive subtitles? Which language edition becomes searchable? Which report is summarized for officials? Which minority language receives no interface or moderation support? Translation resources are finite, but repeated asymmetry can create a durable political filter. descriptive assertion
  3. CL-46-01.03 Access rule: readers should be able to distinguish absence caused by resource limits from absence caused by policy, censorship, or deliberate narrative selection. descriptive assertion
Evidence + editorial analysis

Meaning can change through omission, framing, and register

CL-46-02
As of
Next review
Claim kind
global synthesis

Section summary: Translation requires choices about tone, idiom, ambiguity, institutional names, legal concepts, and cultural references. A translator may clarify one meaning and lose another. Censorship occurs when a material change is imposed or concealed: agency disappears from a sentence, an atrocity becomes an incident, a dissident becomes a criminal, or a skeptical

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
  • SRC-C2PA-SPEC-2-4 C2PA Specifications 2.4 · Official technical standard
Inspect 8 inventoried assertions
  1. CL-46-02.01 Translation requires choices about tone, idiom, ambiguity, institutional names, legal concepts, and cultural references. A translator may clarify one meaning and lose another. Censorship occurs when a material change is imposed or concealed: agency disappears from a sentence, an atrocity becomes an incident, a dissident becomes a criminal, or a skeptical statement becomes certainty. inference or risk
  2. CL-46-02.02 Subtitles and headlines are especially powerful because space is limited and most viewers will never consult the source. Diplomatic and official translation can exploit ambiguity to present different commitments to different audiences. A transparent edition should record omissions, summarize constraints, and preserve difficult terms rather than silently resolving every dispute. descriptive assertion
  3. CL-46-02.03 Element Integrity practice descriptive assertion
  4. CL-46-02.04 Original text Preserve and link the source edition descriptive assertion
  5. CL-46-02.05 Translator Identify role, expertise, and conflicts where safe descriptive assertion
  6. CL-46-02.06 Omissions Mark cuts, summaries, and unavailable passages descriptive assertion
  7. CL-46-02.07 Contested terms Expose alternatives and explain the choice descriptive assertion
  8. CL-46-02.08 Revision Keep version history and correction notes descriptive assertion
Evidence + editorial analysis

Machine translation can scale both access and censorship

CL-46-03
As of
Next review
Claim kind
global synthesis

Section summary: AI can make material available across languages at unprecedented speed. It can also reproduce training-data imbalances, state-mandated exclusions, safety filters, and language-conditioned political bias. Two users may receive different translations because of region, account, model version, or hidden policy, yet both outputs appear fluent and authoritative.

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-C2PA-SPEC-2-4 C2PA Specifications 2.4 · Official technical standard
  • SRC-UNESCO-AI-ETHICS UNESCO Recommendation on the Ethics of Artificial Intelligence · Primary law or court record
Inspect 6 inventoried assertions
  1. CL-46-03.01 AI can make material available across languages at unprecedented speed. It can also reproduce training-data imbalances, state-mandated exclusions, safety filters, and language-conditioned political bias. Two users may receive different translations because of region, account, model version, or hidden policy, yet both outputs appear fluent and authoritative. inference or risk
  2. CL-46-03.02 Quality metrics often emphasize semantic similarity or readability and may miss political omissions, altered agency, or culturally specific meaning. Audits should use matched documents across language directions, scripts, regions, and providers. High-risk material should retain source-aligned segments, uncertainty, and human review rather than presenting one synthetic rendering as definitive. inference or risk
  3. CL-46-03.03 Source Identify exact text, edition, and language. descriptive assertion
  4. CL-46-03.04 Transform Record model, prompt, settings, and human edits. descriptive assertion
  5. CL-46-03.05 Compare Test omissions, entities, agency, stance, and uncertainty. descriptive assertion
  6. CL-46-03.06 Publish Expose provenance and a route to correction. descriptive assertion
Site policy proposal

A transparent multilingual information commons

CL-46-04
As of
Next review
Claim kind
policy proposal

Section summary: Preserve and publish source-language originals whenever law and safety permit. Label abridgment, localization, safety filtering, and politically required changes. Support multiple independent translations and side-by-side comparison for contested material. Protect translators, diaspora journalists, fansub communities, and crisis-language workers from

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-C2PA-SPEC-2-4 C2PA Specifications 2.4 · Official technical standard
  • SRC-UNESCO-AI-ETHICS UNESCO Recommendation on the Ethics of Artificial Intelligence · Primary law or court record
  • SRC-CLEAR-GLOBAL CLEAR Global — language access and crisis communication · Professional or project methodology
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 9 inventoried assertions
  1. CL-46-04.01 A global information space is not open merely because a source technically exists somewhere. It is open when people can reach, understand, compare, and challenge the translations through which that source enters their world. proposal
  2. CL-46-04.02 Preserve and publish source-language originals whenever law and safety permit. proposal
  3. CL-46-04.03 Label abridgment, localization, safety filtering, and politically required changes. proposal
  4. CL-46-04.04 Support multiple independent translations and side-by-side comparison for contested material. proposal
  5. CL-46-04.05 Protect translators, diaspora journalists, fansub communities, and crisis-language workers from retaliation and financial exclusion. proposal
  6. CL-46-04.06 Audit machine translation for differential omission, political framing, minority-language quality, and region-specific outputs. proposal
  7. CL-46-04.07 Use versioned provenance and cryptographic integrity for source and translation artifacts without forcing the translator’s legal identity into public metadata. proposal
  8. CL-46-04.08 Fund low-resource languages, terminology databases, and community review. proposal
  9. CL-46-04.09 Teach readers that fluency is not proof of fidelity. proposal
4 evidence units
Evidence + editorial analysis

Censorship can act at every layer of the network

CL-47-01
As of
Next review
Claim kind
global synthesis

Section summary: Information access depends on more than a website. A state or company can cut physical connectivity, manipulate routes, poison naming, inspect transport metadata, block protocols, remove an app, pressure a host, require identity, deny payments, or prevent a user from receiving an account code. Defending only the content layer leaves the rest of the stack

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 8 inventoried assertions
  1. CL-47-01.01 Information access depends on more than a website. A state or company can cut physical connectivity, manipulate routes, poison naming, inspect transport metadata, block protocols, remove an app, pressure a host, require identity, deny payments, or prevent a user from receiving an account code. Defending only the content layer leaves the rest of the stack available for control. descriptive assertion
  2. CL-47-01.02 Layered intervention also enables plausible deniability. Throttling can resemble poor performance. DNS manipulation can look like a missing site. App-store removal can be described as policy enforcement. Payment loss can appear commercial. Measurement and transparency are therefore essential to distinguish technical failure from deliberate interference. descriptive assertion
  3. CL-47-01.03 Layer Control Resilience descriptive assertion
  4. CL-47-01.04 Physical and routing Shutdowns, route withdrawal, chokepoints Diverse links, exchanges, mesh, and fallback paths descriptive assertion
  5. CL-47-01.05 Naming and transport DNS poisoning, SNI filtering, protocol blocking Encrypted naming, metadata protection, adaptable transports descriptive assertion
  6. CL-47-01.06 Hosting and content Takedown, seizure, deletion Replication, federation, content addressing, archives descriptive assertion
  7. CL-47-01.07 Application and identity Delisting, real-name gates, account denial Open clients, portability, anonymous credentials descriptive assertion
  8. CL-47-01.08 Economic Payment and advertising exclusion Diverse lawful funding and interoperable payment paths descriptive assertion
Evidence + editorial analysis

Resilience comes from plurality and graceful failure

CL-47-02
As of
Next review
Claim kind
global synthesis

Section summary: A centralized service may be efficient, easy to moderate, and simple to use, but one legal order or technical failure can affect everyone. A decentralized network distributes control but may struggle with discovery, abuse response, governance, and inconsistent performance. Federation offers independent operators with shared protocols, yet dominant instances

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-TOR-CENSORSHIP Tor Project — censorship circumvention support · Professional or project methodology
  • SRC-ACCESS-NOW-KEEPITON Access Now — #KeepItOn coalition · Civil-society or secondary analysis
Inspect 3 inventoried assertions
  1. CL-47-02.01 A centralized service may be efficient, easy to moderate, and simple to use, but one legal order or technical failure can affect everyone. A decentralized network distributes control but may struggle with discovery, abuse response, governance, and inconsistent performance. Federation offers independent operators with shared protocols, yet dominant instances or relays can still emerge. inference or risk
  2. CL-47-02.02 The durable design principle is not maximal decentralization at every layer. It is avoiding irreversible dependence. Users should be able to export identity and content, choose providers and clients, communicate across implementations, retain local copies, and fall back to alternative routes. Systems should fail into reduced service rather than total silence. descriptive assertion
  3. CL-47-02.03 Resilience rule: open standards plus practical exit are more important than a platform’s promise that it will remain benevolent. descriptive assertion
Evidence + editorial analysis

Security and abuse do not require a universal observer

CL-47-03
As of
Next review
Claim kind
global synthesis

Section summary: Open networks face malware, fraud, spam, harassment, exploitation, and coordinated abuse. Decentralization can make rapid removal and attribution harder. That does not mean the only alternative is universal identity, weakened encryption, or scanning every private device. Safety can operate through local filters, rate limits, reputation scoped to a service,

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 6 inventoried assertions
  1. CL-47-03.01 Open networks face malware, fraud, spam, harassment, exploitation, and coordinated abuse. Decentralization can make rapid removal and attribution harder. That does not mean the only alternative is universal identity, weakened encryption, or scanning every private device. Safety can operate through local filters, rate limits, reputation scoped to a service, user-chosen trust lists, targeted legal process, and privacy-preserving threat checks. descriptive assertion
  2. CL-47-03.02 Client-side scanning is structurally dangerous because it makes endpoint inspection a normal condition of secure communication. The target database or classifier can expand while the population-wide mechanism remains. A better approach asks which harm must be prevented, which layer can address it with the least collateral observation, and how errors can be appealed. inference or risk
  3. CL-47-03.03 Prevent locally Use user-controlled filters and device protections. descriptive assertion
  4. CL-47-03.04 Coordinate narrowly Share threat signals without raw activity histories. descriptive assertion
  5. CL-47-03.05 Investigate specifically Use evidence and lawful targeted process. descriptive assertion
  6. CL-47-03.06 Measure openly Publish effectiveness, errors, and collateral impact. descriptive assertion
Site policy proposal

Design principles for an open global network

CL-47-04
As of
Next review
Claim kind
policy proposal

Section summary: Keep protocols open, interoperable, implementable, and resistant to control by one vendor or state. Encrypt content and reduce exposed metadata, naming, and identity at each layer. Support provider choice, account and social-graph portability, local export, and independent clients. Distribute hosting, archives, naming, relays, and physical routes across

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
Inspect 9 inventoried assertions
  1. CL-47-04.01 An uncensored internet will never be free of rules, conflict, or harmful use. Its defining property is that no single rule-maker can quietly turn the global network into a universal permission system for human inquiry. proposal
  2. CL-47-04.02 Keep protocols open, interoperable, implementable, and resistant to control by one vendor or state. proposal
  3. CL-47-04.03 Encrypt content and reduce exposed metadata, naming, and identity at each layer. proposal
  4. CL-47-04.04 Support provider choice, account and social-graph portability, local export, and independent clients. proposal
  5. CL-47-04.05 Distribute hosting, archives, naming, relays, and physical routes across organizations and jurisdictions. proposal
  6. CL-47-04.06 Build anti-censorship adaptability without promising one protocol will remain undetectable forever. proposal
  7. CL-47-04.07 Address abuse with the narrowest effective layer, user controls, targeted process, and transparent governance. proposal
  8. CL-47-04.08 Protect security research, network measurement, encryption, and lawful circumvention tools. proposal
  9. CL-47-04.09 Require shutdowns, blocking, and surveillance orders to be lawful, specific, reviewable, time-limited, and publicly accounted for. proposal
4 evidence units
Evidence + editorial analysis

Operational secrecy has a legitimate and narrow core

CL-48-01
As of
Next review
Claim kind
global synthesis

Section summary: Military operations depend on surprise, secure communications, protected identities, and control of vulnerabilities. Real-time disclosure of unit locations, planned routes, intelligence methods, defensive gaps, or hostage-rescue details can create a direct and foreseeable risk to life. A free society can recognize these limits without treating every

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-SCOTUS-PENTAGON-PAPERS New York Times Co. v. United States — bound U.S. Reports · Primary law or court record
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 3 inventoried assertions
  1. CL-48-01.01 Military operations depend on surprise, secure communications, protected identities, and control of vulnerabilities. Real-time disclosure of unit locations, planned routes, intelligence methods, defensive gaps, or hostage-rescue details can create a direct and foreseeable risk to life. A free society can recognize these limits without treating every war-related fact as public on demand. inference or risk
  2. CL-48-01.02 The justification is strongest when the information is specific, the threat is imminent, the secrecy period is short, and no narrower remedy is available. Delayed publication, location blurring, source protection, temporary embargo, or removal of one operational detail may protect the mission without suppressing the broader investigation. inference or risk
  3. CL-48-01.03 Operational test: identify the exact information, adversary advantage, likely harm, time window, and less restrictive alternative. “National security” alone is not an explanation. descriptive assertion
Evidence + editorial analysis

The censorship ratchet begins when military harm becomes political harm

CL-48-02
As of
Next review
Claim kind
global synthesis

Section summary: Wartime institutions often expand secrecy from operations to narratives. Casualty figures, civilian harm, corruption, torture, failed plans, dissent, or criticism may be restricted because disclosure damages morale, embarrasses leaders, or weakens support. Those are political consequences, not automatically operational threats. Vague offenses such as

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
  • SRC-ACCESS-NOW-KEEPITON Access Now — #KeepItOn coalition · Civil-society or secondary analysis
Inspect 8 inventoried assertions
  1. CL-48-02.01 Wartime institutions often expand secrecy from operations to narratives. Casualty figures, civilian harm, corruption, torture, failed plans, dissent, or criticism may be restricted because disclosure damages morale, embarrasses leaders, or weakens support. Those are political consequences, not automatically operational threats. inference or risk
  2. CL-48-02.02 Vague offenses such as discrediting the armed forces, spreading false information, aiding the enemy, or undermining unity can make official statements the only legally safe account. Journalists and citizens then cannot test whether the state’s claims are accurate. Secrecy that prevents accountability may protect an administration while weakening the society it claims to defend. inference or risk
  3. CL-48-02.03 Question Operational secrecy Political censorship descriptive assertion
  4. CL-48-02.04 Target Specific capability, location, or plan Criticism, evidence, or dissent descriptive assertion
  5. CL-48-02.05 Time Ends when the risk passes Indefinite or retroactive inference or risk
  6. CL-48-02.06 Evidence Concrete causal pathway to harm Embarrassment, morale, or disagreement descriptive assertion
  7. CL-48-02.07 Review Independent and adversarial where possible Executive or military self-approval descriptive assertion
  8. CL-48-02.08 Record Preserved for later release Destroyed, altered, or permanently sealed descriptive assertion
Evidence + editorial analysis

Digital war makes total information control both tempting and destructive

CL-48-03
As of
Next review
Claim kind
global synthesis

Section summary: Satellites, smartphones, social media, open-source intelligence, drones, and AI make operational information travel quickly. States may respond with network shutdowns, platform blocks, device seizures, or broad bans on imagery. These measures can also disable emergency services, humanitarian coordination, independent reporting, evidence preservation,

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-ACCESS-NOW-KEEPITON Access Now — #KeepItOn coalition · Civil-society or secondary analysis
  • SRC-SYRIAN-ARCHIVE Syrian Archive — preservation and verification programme · Professional or project methodology
Inspect 6 inventoried assertions
  1. CL-48-03.01 Satellites, smartphones, social media, open-source intelligence, drones, and AI make operational information travel quickly. States may respond with network shutdowns, platform blocks, device seizures, or broad bans on imagery. These measures can also disable emergency services, humanitarian coordination, independent reporting, evidence preservation, education, banking, and family contact. inference or risk
  2. CL-48-03.02 AI complicates verification by generating persuasive synthetic media and accelerating analysis of public data. The answer is not one official information monopoly. Newsrooms, platforms, militaries, and investigators need provenance, source comparison, careful delay of operational details, correction, and transparent uncertainty. Independent archives should preserve originals even when immediate public release is unsafe. descriptive assertion
  3. CL-48-03.03 Acquire Protect sources and preserve exact evidence. descriptive assertion
  4. CL-48-03.04 Assess Separate operational danger from political sensitivity. inference or risk
  5. CL-48-03.05 Publish carefully Delay or redact only what creates concrete harm. descriptive assertion
  6. CL-48-03.06 Release later Review secrecy and restore the historical record. descriptive assertion
Site policy proposal

A wartime information-rights framework

CL-48-04
As of
Next review
Claim kind
policy proposal

Section summary: Define protected operational information narrowly and require a documented, concrete harm assessment. Use delay, redaction, pooling, or location protection before broader publication bans. Keep independent courts, legislatures, inspectors, and news organizations able to review secrecy claims. Protect criticism of policy, reporting of civilian harm,

Support scope: The exact supplied report and the listed visible sources support the section’s factual or conceptual foundation. Comparative framing, worldwide implications, and proposed safeguards remain CognitiveLiberties.com editorial synthesis.

Limit: The source package was inspected and preserved, but these external locators were not independently reopened in this integration round. Current law, deployment, statistics, provider behavior, and technical capability require dated verification before high-stakes reliance.

Conflict or tension: Safety, authenticity, fraud prevention, public order, operational security, and service integrity may support narrower interventions. The section evaluates when those interventions become opaque, durable, identity-linked, or broader than the demonstrated harm.

Mapped sources
  • SRC-ACCESS-NOW-KEEPITON Access Now — #KeepItOn coalition · Civil-society or secondary analysis
  • SRC-SYRIAN-ARCHIVE Syrian Archive — preservation and verification programme · Professional or project methodology
  • SRC-SCOTUS-PENTAGON-PAPERS New York Times Co. v. United States — bound U.S. Reports · Primary law or court record
  • SRC-OHCHR-GC34 UN Human Rights Committee — General Comment No. 34 on Article 19 · Primary law or court record
Inspect 9 inventoried assertions
  1. CL-48-04.01 The purpose of wartime secrecy should be to protect people and operations, not to make a government’s conduct unknowable. The society that emerges from conflict needs an intact press, historical record, and capacity to judge what was done in its name. proposal
  2. CL-48-04.02 Define protected operational information narrowly and require a documented, concrete harm assessment. proposal
  3. CL-48-04.03 Use delay, redaction, pooling, or location protection before broader publication bans. proposal
  4. CL-48-04.04 Keep independent courts, legislatures, inspectors, and news organizations able to review secrecy claims. proposal
  5. CL-48-04.05 Protect criticism of policy, reporting of civilian harm, whistleblowing, and evidence of unlawful conduct. proposal
  6. CL-48-04.06 Restrict shutdowns to the smallest area and time, publish legal authority, and assess humanitarian impact. proposal
  7. CL-48-04.07 Preserve all restricted records with fixity, custody, and automatic declassification or review dates. proposal
  8. CL-48-04.08 Prohibit emergency powers from becoming permanent identity, surveillance, or censorship infrastructure after the conflict. proposal
  9. CL-48-04.09 Provide later notice, appeal, correction, and remedy for wrongful suppression. proposal
Public source map

Source registry

Source class and retrieval state are part of the evidence. A first-party product document can establish what a provider publishes, but it does not independently prove production behavior. A preprint can report original research, but its scope and review status remain relevant.

SRC-ACADEMIC-FREEDOM-INDEX

Academic Freedom Index

Publisher
Academic Freedom Index project
Class
Original peer-reviewed research
Jurisdiction or scope
Global comparative research
Retrieval
Not independently reopened in this release
Next review

Authority: Cross-national measurement programme covering freedom to research, teach, exchange, and communicate academic findings.

Limit: Composite indicators depend on methodology, expert coding, country coverage, uncertainty, and the edition being consulted.

SRC-ACCESS-NOW-KEEPITON

Access Now — #KeepItOn coalition

Publisher
Access Now
Class
Civil-society or secondary analysis
Jurisdiction or scope
Worldwide internet shutdown monitoring
Retrieval
Not independently reopened in this release
Next review

Authority: Civil-society documentation and advocacy concerning internet shutdowns, network disruptions, and their human-rights consequences.

Limit: Event counts and attribution depend on measurement access, local reporting, definitions, and the period under review.

SRC-ACLU-BORDER-ZONE

ACLU — the Constitution in the 100-mile border zone

Publisher
American Civil Liberties Union
Class
Civil-society or secondary analysis
Jurisdiction or scope
United States
Retrieval
Not independently reopened in this release
Next review

Authority: Civil-liberties guidance describing border-zone authorities and contested constitutional limits.

Limit: General guidance is not a substitute for current statutes, regulations, and circuit-specific case law.

SRC-ALA-LMS-PRIVACY

American Library Association — privacy guidelines for library management systems

Publisher
American Library Association
Class
Professional or project methodology
Jurisdiction or scope
United States library practice
Retrieval
Not independently reopened in this release
Next review

Authority: Professional privacy guidance for minimizing and protecting patron records in library systems.

Limit: Implementation and vendor contracts must be independently audited.

SRC-ALA-MINORS-ONLINE

American Library Association — minors and online activity

Publisher
American Library Association
Class
Professional or project methodology
Jurisdiction or scope
United States library ethics
Retrieval
Not independently reopened in this release
Next review

Authority: Professional interpretation supporting minors’ intellectual freedom and individualized responses to misuse.

Limit: Professional principles are not binding law and library obligations vary.

SRC-ARXIV-BUGS-CSS-2021

Bugs in our Pockets — The Risks of Client-Side Scanning

Publisher
Original multi-author technical report
Class
Original preprint or working paper
Jurisdiction or scope
Technical research
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Technical and security analysis of client-side scanning and its evasion, abuse, and expansion risks.

Limit: It is a critical analysis rather than a comparative field evaluation of every possible implementation.

SRC-ARXIV-LLM-PERSONALITY-2024

Large Language Models Can Infer Personality from Free-Form User Interactions

Publisher
Original research paper
Class
Original preprint or working paper
Jurisdiction or scope
AI research
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Study evaluating personality-trait inference from free-form interactions.

Limit: Reported performance is study- and dataset-specific and does not justify treating inferred traits as ground truth.

SRC-ARXIV-ORBENCH-2024

OR-Bench — over-refusal benchmark

Publisher
Original research paper
Class
Original peer-reviewed research
Jurisdiction or scope
AI research
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Large benchmark designed to evaluate over-refusal on superficially sensitive but benign prompts.

Limit: Model results are version-specific and cannot establish the behavior of every proprietary deployment.

SRC-ARXIV-QUERY-INTENT-2020

Deep Search Query Intent Understanding

Publisher
Original search-retrieval research
Class
Original preprint or working paper
Jurisdiction or scope
Information retrieval research
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Original research describing character- and word-level models for scalable query-intent prediction.

Limit: The paper documents particular systems and experiments, not the complete current pipelines of all search providers.

SRC-ARXIV-RAG-2020

Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks

Publisher
Original research paper
Class
Original peer-reviewed research
Jurisdiction or scope
AI research
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Foundational paper combining parametric generation with a retrieved non-parametric memory.

Limit: Modern production generative search systems can differ substantially from the original architecture.

SRC-ARXIV-XSTEST-2023

XSTest — exaggerated safety behavior benchmark

Publisher
Original research paper
Class
Original peer-reviewed research
Jurisdiction or scope
AI research
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Benchmark containing safe and unsafe contrast prompts to measure false refusals and decision-boundary calibration.

Limit: Benchmark results depend on model version, prompt format, evaluator, and deployment context.

SRC-BOSTROM-INFORMATION-HAZARDS

Nick Bostrom — Information Hazards

Publisher
Original philosophical paper
Class
Original preprint or working paper
Jurisdiction or scope
Philosophy and risk scholarship
Retrieval
Not independently reopened in this release
Source date
Next review

Authority: Conceptual taxonomy of risks arising from the dissemination or possession of information.

Limit: A philosophical taxonomy does not establish the probability or proper control of any specific hazard.

SRC-BROOKINGS-ISRAEL-CONTACT-TRACING

Brookings — technosurveillance mission creep in Israel’s COVID-19 response

Publisher
Brookings Institution
Class
Civil-society or secondary analysis
Jurisdiction or scope
Israel / comparative policy
Retrieval
Not independently reopened in this release
Source date
Next review

Authority: Policy analysis of repurposed security surveillance in a public-health setting.

Limit: Secondary analysis should be checked against the underlying legal orders and implementation record.

SRC-C2PA-SPEC-2-4

C2PA Specifications 2.4

Publisher
Coalition for Content Provenance and Authenticity
Class
Official technical standard
Jurisdiction or scope
Digital media provenance standard
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Current specification suite for certifying source and history information associated with media content.

Limit: C2PA conveys provenance assertions under a trust model; it does not independently prove that content is true or guarantee search visibility.

SRC-CAMBRIDGE-SECURITIZATION

Cambridge overview of securitization theory

Publisher
Cambridge University Press
Class
Civil-society or secondary analysis
Jurisdiction or scope
International-relations scholarship
Retrieval
Not independently reopened in this release
Next review

Authority: Scholarly overview of security framing and emergency politics.

Limit: The theory is an analytical framework and does not prove bad faith in a particular policy process.

SRC-CITIZEN-LAB-CENSORSHIP

Citizen Lab — internet censorship research

Publisher
Citizen Lab, University of Toronto
Class
Civil-society or secondary analysis
Jurisdiction or scope
Global technical investigations
Retrieval
Not independently reopened in this release
Next review

Authority: Technical investigations into filtering, surveillance, spyware, and commercial censorship systems.

Limit: Individual studies have bounded datasets, periods, and attribution confidence.

SRC-CLEAR-GLOBAL

CLEAR Global — language access and crisis communication

Publisher
CLEAR Global
Class
Professional or project methodology
Jurisdiction or scope
Worldwide language access
Retrieval
Not independently reopened in this release
Next review

Authority: Project practice and research concerning translation, language inclusion, and access to critical information across linguistic communities.

Limit: Organizational materials demonstrate methods and needs, not a comprehensive account of all translation censorship or machine-translation bias.

SRC-COLORADO-HB24-1058

Colorado HB 24-1058 — signed act

Publisher
Colorado General Assembly
Class
Primary law or court record
Jurisdiction or scope
Colorado, United States
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Signed legislation expanding Colorado privacy protections to biological data including neural data.

Limit: The statutory definition and coverage should not be generalized to all inferred mental data or all jurisdictions.

SRC-DATA-PROVENANCE-INITIATIVE

Data Provenance Initiative — dataset documentation and audit research

Publisher
Data Provenance Initiative
Class
Professional or project methodology
Jurisdiction or scope
Global AI dataset research
Retrieval
Not independently reopened in this release
Next review

Authority: Project documentation and research focused on tracing, documenting, and evaluating the data used to build AI systems.

Limit: Coverage depends on available datasets and disclosures; provenance documentation does not prove that every model uses the described data.

SRC-DOJ-OIG-FBI-NSL

U.S. Department of Justice Inspector General — FBI and National Security Letter reports

Publisher
U.S. Department of Justice Office of the Inspector General
Class
Official oversight record
Jurisdiction or scope
United States
Retrieval
Reviewed online on 2026-09-02
Next review

Authority: Official index of inspector-general reviews documenting implementation, compliance, and control failures.

Limit: The index contains multiple reports with different periods and findings; a specific proposition should identify the relevant report.

SRC-DUKE-DATA-BROKERS

Duke report on data brokers and sensitive data on U.S. individuals

Publisher
Duke Sanford School of Public Policy
Class
Original preprint or working paper
Jurisdiction or scope
United States
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Original policy research documenting market availability of sensitive personal data.

Limit: Broker offerings and practices change; the report does not establish how every buyer uses the data.

SRC-DUKE-JUDICATURE-COGNITIVE-LIBERTY

Duke Judicature — The Battle for Your Brain

Publisher
Judicature / Duke Law
Class
Expert interview or commentary
Jurisdiction or scope
United States legal commentary
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Expert legal argument for protecting brain data and cognitive liberty.

Limit: Expert commentary is not a judicial holding or enacted legal rule.

SRC-DUKE-LAW-COGNITIVE-LIBERTY

Duke Law — cognitive liberty Q&A

Publisher
Duke Law Magazine
Class
Expert interview or commentary
Jurisdiction or scope
United States legal commentary
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Expert discussion of mental privacy, cognitive liberty, and emerging technology.

Limit: Interview framing should not be presented as settled doctrine or empirical proof.

SRC-EFF-KEYWORD-WARRANTS

EFF — keyword search warrants

Publisher
Electronic Frontier Foundation
Class
Civil-society or secondary analysis
Jurisdiction or scope
United States
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Civil-liberties analysis of reverse-keyword warrants and their Fourth Amendment implications.

Limit: Advocacy interpretation should be checked against the warrant, record, and controlling court opinions.

SRC-EFF-PATRIOT-ACT

EFF analysis of the USA PATRIOT Act

Publisher
Electronic Frontier Foundation
Class
Civil-society or secondary analysis
Jurisdiction or scope
United States
Retrieval
Not independently reopened in this release
Source date
Next review

Authority: Civil-liberties analysis of the Act’s surveillance provisions and risks.

Limit: Advocacy analysis should be read alongside statutes, cases, and official oversight findings.

SRC-FIRST-MONDAY-POLITICAL-HUMOR

First Monday — political satire and coded online expression

Publisher
First Monday
Class
Original peer-reviewed research
Jurisdiction or scope
Digital political expression case study
Retrieval
Not independently reopened in this release
Next review

Authority: Scholarly analysis of online political humor, coded language, and cultural adaptation under restrictive information conditions.

Limit: A bounded case study should not be generalized to all humor, languages, platforms, or political systems without additional evidence.

SRC-FREEDOM-HOUSE-FOTN

Freedom House — Freedom on the Net

Publisher
Freedom House
Class
Civil-society or secondary analysis
Jurisdiction or scope
Global comparative research
Retrieval
Not independently reopened in this release
Next review

Authority: Country research on obstacles to access, content limits, and violations of user rights.

Limit: Scores reflect the project methodology and should not be treated as infallible or timeless.

SRC-FREEDOM-HOUSE-TRANSNATIONAL

Freedom House — transnational repression research

Publisher
Freedom House
Class
Civil-society or secondary analysis
Jurisdiction or scope
Cross-border repression
Retrieval
Not independently reopened in this release
Next review

Authority: Research on states pursuing critics, journalists, and diaspora communities beyond territorial borders.

Limit: Incident coverage and attribution depend on available evidence and reporting.

SRC-FTC-BIOMETRIC-POLICY-2023

FTC policy statement on biometric information and Section 5

Publisher
Federal Trade Commission
Class
Official oversight record
Jurisdiction or scope
United States
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Official Commission policy statement describing unfair or deceptive biometric-information practices under Section 5.

Limit: A policy statement explains enforcement views; it is not itself a statute or final adjudication in every application.

SRC-GNI-PRINCIPLES

Global Network Initiative — Principles on Freedom of Expression and Privacy

Publisher
Global Network Initiative
Class
Professional or project methodology
Jurisdiction or scope
Global business and human-rights framework
Retrieval
Not independently reopened in this release
Next review

Authority: Multi-stakeholder principles addressing company responsibilities when government demands affect expression and privacy.

Limit: Voluntary principles require implementation and independent evidence; membership or stated commitment does not prove compliance in every case.

SRC-GOOGLE-AI-OPTIMIZATION

Google guide to optimizing for generative AI features

Publisher
Google Search Central
Class
First-party platform guidance
Jurisdiction or scope
Google Search
Retrieval
Reviewed online on 2026-09-02
Next review

Authority: Current first-party guidance stating that core SEO practices remain relevant to Google’s generative search features.

Limit: The guide describes Google’s published practices and does not guarantee citation, traffic, or ranking.

SRC-GOOGLE-RATER-GUIDELINES

Google Search Quality Rater Guidelines

Publisher
Google
Class
First-party platform guidance
Jurisdiction or scope
Google Search quality evaluation
Retrieval
Not independently reopened in this release
Next review

Authority: First-party evaluator guidance used to assess search quality examples and concepts.

Limit: Rater guidelines are not a complete list of ranking signals and should not be represented as a direct scoring formula.

SRC-GOOGLE-RECSYS-TYPES

Google for Developers — recommendation system overview

Publisher
Google
Class
First-party platform guidance
Jurisdiction or scope
General recommendation-system engineering
Retrieval
Not independently reopened in this release
Next review

Authority: First-party educational documentation describing common recommendation architectures and stages such as candidate generation and scoring.

Limit: Educational documentation does not reveal every production ranking signal, policy intervention, or real-world suppression decision.

SRC-GOOGLE-SEARCH-ESSENTIALS

Google Search Essentials

Publisher
Google Search Central
Class
First-party platform guidance
Jurisdiction or scope
Google Search
Retrieval
Reviewed online on 2026-09-02
Next review

Authority: First-party requirements and policy guidance for eligibility and performance in Google Search.

Limit: Following the guidance does not guarantee crawling, indexing, ranking, or inclusion in generative features.

SRC-HARVARD-PENNEY-2016

Harvard Berkman Klein Center — Jon Penney research event

Publisher
Berkman Klein Center for Internet & Society
Class
Civil-society or secondary analysis
Jurisdiction or scope
Academic event record
Retrieval
Not independently reopened in this release
Source date
Next review

Authority: Institutional event page summarizing research on surveillance and Wikipedia use.

Limit: An event summary is not a substitute for the full methods and results paper.

SRC-IEEE-7000-FAMILY

IEEE 7000 standards family overview

Publisher
IEEE Standards Association
Class
Official technical standard
Jurisdiction or scope
Technology standards
Retrieval
Not independently reopened in this release
Next review

Authority: Standards-family overview for value-sensitive and accountable autonomous-system design.

Limit: The overview does not establish compliance by any particular deployed system.

SRC-IENCA-ANDORNO-NEURORIGHTS

On Neurorights

Publisher
Frontiers in Human Neuroscience / PubMed Central
Class
Original peer-reviewed research
Jurisdiction or scope
Neuroethics and human-rights scholarship
Retrieval
Retrieval blocked or unavailable on 2026-09-02
Source date
Next review

Authority: Peer-reviewed conceptual analysis of neurorights and cognitive liberty.

Limit: Direct retrieval was interrupted during this review; the article is scholarly argument, not binding law.

SRC-ILLINOIS-PA-103-0100

Illinois Public Act 103-0100

Publisher
Illinois General Assembly
Class
Primary law or court record
Jurisdiction or scope
Illinois, United States
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Official state law connecting grant eligibility to anti-book-ban or Library Bill of Rights policies.

Limit: The act addresses Illinois library grant eligibility; it does not regulate all private or digital content systems.

SRC-INTERNET-SOCIETY-FRAGMENTATION

Internet Society — internet fragmentation resources

Publisher
Internet Society
Class
Professional or project methodology
Jurisdiction or scope
Global internet architecture
Retrieval
Not independently reopened in this release
Next review

Authority: Technical and policy analysis of fragmentation, interoperability, and network-level control.

Limit: The collection explains risks and methods; it is not a live census of every country.

SRC-IPFS-DOCS

IPFS documentation — content-addressed and peer-to-peer storage

Publisher
InterPlanetary File System project
Class
First-party platform guidance
Jurisdiction or scope
Global distributed-systems architecture
Retrieval
Not independently reopened in this release
Next review

Authority: First-party technical explanation of content addressing, peer-to-peer distribution, and resilient retrieval in IPFS.

Limit: Documentation explains architecture, not guaranteed persistence, legality, authenticity, availability, or safe governance in every deployment.

SRC-MARTHEWS-TUCKER-SEARCH-CHILL

Marthews and Tucker — Government Surveillance and Internet Search Behavior

Publisher
Original working paper
Class
Original preprint or working paper
Jurisdiction or scope
Surveillance and search-behavior research
Retrieval
Retrieval blocked or unavailable on 2026-09-02
Source date
Next review

Authority: Empirical analysis associated with changes in sensitive search behavior after surveillance disclosures.

Limit: Direct SSRN retrieval was blocked; the public synthesis should preserve causal and measurement caveats.

SRC-NATURE-SEMANTIC-DECODER-2023

Semantic reconstruction of continuous language from non-invasive brain recordings

Publisher
Nature Neuroscience
Class
Original peer-reviewed research
Jurisdiction or scope
Neuroscience research
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Original study demonstrating semantic reconstruction under controlled fMRI conditions with subject cooperation and individualized training.

Limit: The study does not demonstrate unrestricted remote mind reading; cooperation, training, equipment, and experimental conditions are material limits.

SRC-NCMEC-CYBERTIPLINE-DATA

NCMEC CyberTipline data

Publisher
National Center for Missing & Exploited Children
Class
Official oversight record
Jurisdiction or scope
United States
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: First-party program data describing reports of suspected online child sexual exploitation.

Limit: Reports are not counts of unique verified crimes or victims and should not be treated as adjudicated cases.

SRC-OHCHR-A76-380

UN Special Rapporteur report A/76/380 — Freedom of thought

Publisher
Office of the UN High Commissioner for Human Rights
Class
Official oversight record
Jurisdiction or scope
International
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Thematic report examining the scope and contemporary vulnerability of freedom of thought.

Limit: A Special Rapporteur report is expert UN analysis, not binding adjudication.

SRC-OHCHR-ANONYMITY-ENCRYPTION-2015

UN Special Rapporteur — encryption and anonymity in digital communications

Publisher
United Nations Digital Library
Class
Official oversight record
Jurisdiction or scope
Global
Retrieval
Not independently reopened in this release
Next review

Authority: Official thematic report addressing encryption, anonymity, privacy, and freedom of expression.

Limit: The report is normative and global; national law and technical deployment differ.

SRC-OHCHR-CRC

Convention on the Rights of the Child

Publisher
Office of the UN High Commissioner for Human Rights
Class
Primary law or court record
Jurisdiction or scope
International
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Treaty text addressing children’s privacy, expression, information access, thought, conscience, and religion.

Limit: Domestic effect and ratification status vary by jurisdiction.

SRC-OHCHR-GC22

Human Rights Committee General Comment No. 22

Publisher
UN Human Rights Committee
Class
Primary law or court record
Jurisdiction or scope
International
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Authoritative committee interpretation distinguishing unconditional internal freedom from regulable manifestation.

Limit: General comments interpret treaty obligations but do not create a standalone domestic cause of action everywhere.

SRC-OHCHR-GC34

UN Human Rights Committee — General Comment No. 34 on Article 19

Publisher
Office of the UN High Commissioner for Human Rights
Class
Primary law or court record
Jurisdiction or scope
International human-rights framework
Retrieval
Not independently reopened in this release
Next review

Authority: Authoritative interpretation of freedom of opinion and expression under the ICCPR.

Limit: It supplies a legal framework, not a finding that every national implementation complies.

SRC-OHCHR-ICCPR

International Covenant on Civil and Political Rights — Article 18

Publisher
Office of the UN High Commissioner for Human Rights
Class
Primary law or court record
Jurisdiction or scope
International
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Treaty text protecting freedom of thought, conscience, and religion.

Limit: The treaty text does not itself define a complete modern doctrine of cognitive liberty.

SRC-OHCHR-RABAT-PLAN

OHCHR — Rabat Plan of Action

Publisher
Office of the UN High Commissioner for Human Rights
Class
Professional or project methodology
Jurisdiction or scope
Global
Retrieval
Not independently reopened in this release
Next review

Authority: International threshold framework for distinguishing protected expression from incitement.

Limit: It is guidance and must be applied to specific legal facts and local procedures.

SRC-OONI-MEASUREMENT

OONI — open measurement of internet censorship

Publisher
Open Observatory of Network Interference
Class
Professional or project methodology
Jurisdiction or scope
Global network measurement
Retrieval
Not independently reopened in this release
Next review

Authority: Public methodology and measurements for detecting network interference from distributed vantage points.

Limit: Coverage depends on volunteer probes, test design, local conditions, and attribution evidence.

SRC-OSI-OPEN-SOURCE-AI-1-0

Open Source AI Definition 1.0

Publisher
Open Source Initiative
Class
Official technical standard
Jurisdiction or scope
Open-source project standard
Retrieval
Not independently reopened in this release
Source date
Next review

Authority: Project definition describing the freedoms and information required for an AI system to be considered open source.

Limit: It is a project definition, not a statute or universal industry mandate.

SRC-PCLOB-SECTION-215-2014

Privacy and Civil Liberties Oversight Board — Section 215 report

Publisher
Privacy and Civil Liberties Oversight Board
Class
Official oversight record
Jurisdiction or scope
United States
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Independent federal oversight report evaluating legal foundation, value, privacy, and civil-liberties costs of bulk telephone records collection.

Limit: The findings concern a particular surveillance program and do not prove that all bulk or AI monitoring systems have identical value or defects.

SRC-PEN-AMERICA-GLOBAL-CHILLING

PEN America — Global Chilling

Publisher
PEN America
Class
Civil-society or secondary analysis
Jurisdiction or scope
International
Retrieval
Not independently reopened in this release
Source date
Next review

Authority: Civil-society research on surveillance concerns and self-censorship among writers internationally.

Limit: Survey composition and self-reporting limit generalization to all populations.

SRC-PEN-AMERICA-WRITERS-SURVEILLANCE

PEN America — impact of NSA surveillance on writers

Publisher
PEN America
Class
Civil-society or secondary analysis
Jurisdiction or scope
United States
Retrieval
Not independently reopened in this release
Source date
Next review

Authority: Civil-society survey reporting writers’ concerns and self-censorship after surveillance disclosures.

Limit: Survey results reflect respondents and self-reported behavior rather than a general-population causal estimate.

SRC-PENNEY-WIKIPEDIA-CHILL

Jon Penney — Chilling Effects: Online Surveillance and Wikipedia Use

Publisher
Original legal-empirical research
Class
Original peer-reviewed research
Jurisdiction or scope
Surveillance and information-seeking research
Retrieval
Retrieval blocked or unavailable on 2026-09-02
Source date
Next review

Authority: Empirical study associated with post-Snowden changes in views of privacy-sensitive Wikipedia pages.

Limit: Direct SSRN retrieval was blocked; observational design cannot eliminate every alternative explanation.

SRC-RANKING-DIGITAL-RIGHTS

Ranking Digital Rights methodology

Publisher
Ranking Digital Rights
Class
Professional or project methodology
Jurisdiction or scope
Global digital-platform accountability
Retrieval
Not independently reopened in this release
Source date
Next review

Authority: Structured methodology for evaluating governance, freedom of expression, and privacy commitments.

Limit: The method mainly assesses disclosed policies and governance; production behavior requires separate testing.

SRC-RFC-9458-OHTTP

RFC 9458 — Oblivious HTTP

Publisher
Internet Engineering Task Force / RFC Editor
Class
Official technical standard
Jurisdiction or scope
Internet standard
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Standards-track protocol for forwarding encrypted HTTP messages through a relay to reduce linkability between clients and requests.

Limit: Privacy depends on deployment assumptions, payload content, and non-collusion; OHTTP is not complete anonymity by itself.

SRC-RFC-9576-PRIVACY-PASS

RFC 9576 — The Privacy Pass Architecture

Publisher
Internet Engineering Task Force / RFC Editor
Class
Official technical standard
Jurisdiction or scope
Internet architecture
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Architecture for privacy-preserving authorization with unlinkable proofs rather than long-term state.

Limit: The document defines an architecture, not a turnkey age-assurance system or proof that every deployment is unlinkable.

SRC-RFC-9614-PRIVACY-PARTITIONING

RFC 9614 — Partitioning as an Architecture for Privacy

Publisher
Internet Architecture Board / RFC Editor
Class
Official technical standard
Jurisdiction or scope
Internet architecture
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Architectural framework for separating identity, data, and metadata across parties under least-privilege principles.

Limit: The RFC expressly warns that partitioning is not a panacea and requires whole-system analysis.

SRC-RICHARDS-INTELLECTUAL-PRIVACY

Neil Richards — Intellectual Privacy

Publisher
SSRN legal scholarship
Class
Original preprint or working paper
Jurisdiction or scope
United States legal scholarship
Retrieval
Retrieval blocked or unavailable on 2026-09-02
Source date
Next review

Authority: Legal scholarship developing intellectual privacy as a foundation for free thought and expression.

Limit: SSRN retrieval was blocked during this review; the paper is scholarship, not binding law.

SRC-SANTA-CLARA-PRINCIPLES

Santa Clara Principles on Transparency and Accountability in Content Moderation

Publisher
Santa Clara Principles project
Class
Professional or project methodology
Jurisdiction or scope
Global platform governance
Retrieval
Not independently reopened in this release
Next review

Authority: Civil-society principles addressing notice, numbers, and appeals in content moderation.

Limit: Voluntary principles are not binding law and implementation varies.

SRC-SCHOLARS-AT-RISK-2025

Scholars at Risk — Free to Think 2025

Publisher
Scholars at Risk Network
Class
Civil-society or secondary analysis
Jurisdiction or scope
Worldwide academic freedom
Retrieval
Not independently reopened in this release
Source date
Next review

Authority: Civil-society monitoring and case documentation concerning attacks on higher education communities and academic freedom.

Limit: Incident reporting is bounded by documentation access, verification methods, and the report period; it is not a complete global census.

SRC-SCOTUS-CARPENTER-2018

Carpenter v. United States — official opinion

Publisher
Supreme Court of the United States
Class
Primary law or court record
Jurisdiction or scope
United States
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Official opinion holding acquisition of historical cell-site records to be a Fourth Amendment search in the circumstances presented.

Limit: The opinion is expressly narrow and does not decide the constitutional status of search-query or AI-assistant histories.

SRC-SCOTUS-MOODY-2024

Moody v. NetChoice — official opinion

Publisher
Supreme Court of the United States
Class
Primary law or court record
Jurisdiction or scope
United States
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Official opinion addressing platform content moderation, editorial discretion, and facial-challenge analysis.

Limit: The Court remanded for a more service-specific analysis; the opinion does not create universal immunity for every platform action.

SRC-SCOTUS-MURTHY-2024

Murthy v. Missouri — official opinion

Publisher
Supreme Court of the United States
Class
Primary law or court record
Jurisdiction or scope
United States
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Official opinion resolving the case on standing and documenting the disputed government-platform communications record.

Limit: The standing disposition did not finally resolve every alleged jawboning practice on the merits.

SRC-SCOTUS-PENTAGON-PAPERS

New York Times Co. v. United States — bound U.S. Reports

Publisher
U.S. Government Publishing Office
Class
Primary law or court record
Jurisdiction or scope
United States historical case with comparative relevance
Retrieval
Not independently reopened in this release
Source date
Next review

Authority: Primary judicial record concerning prior restraint and publication of classified historical material during war.

Limit: The holding is fact-specific and jurisdiction-specific; it does not eliminate legitimate operational secrecy or govern other legal systems.

SRC-SCOTUS-PICO-1982

Board of Education v. Pico — bound U.S. Reports

Publisher
U.S. Government Publishing Office
Class
Primary law or court record
Jurisdiction or scope
United States
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Official bound-volume record of the school-library book-removal case and its influential plurality reasoning.

Limit: The fractured decision does not establish a single broad rule for every school or digital-information dispute.

SRC-SCOTUS-STANFORD-1965

Stanford v. Texas — bound U.S. Reports

Publisher
U.S. Government Publishing Office
Class
Primary law or court record
Jurisdiction or scope
United States
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Official bound-volume opinion applying heightened particularity concerns where expressive materials are searched or seized.

Limit: The case concerns physical books and papers, not modern reverse-keyword warrants directly.

SRC-SIGNAL-SPECS

Signal — protocol and cryptographic specifications

Publisher
Signal
Class
First-party platform guidance
Jurisdiction or scope
Global secure-messaging architecture
Retrieval
Not independently reopened in this release
Next review

Authority: First-party specifications for cryptographic protocols used to protect message content and support forward secrecy.

Limit: Published protocol design does not by itself prove endpoint security, implementation correctness, metadata minimization, or operational safety.

SRC-STANFORD-PRIO-2017

Prio — Private, Robust, and Scalable Computation of Aggregate Statistics

Publisher
Stanford Applied Cryptography Group
Class
Original peer-reviewed research
Jurisdiction or scope
Technical research
Retrieval
Reviewed online on 2026-09-02
Source date
Next review

Authority: Original research on privacy-preserving aggregate statistics using multiple servers and secret-shared proofs.

Limit: Security relies on system assumptions including at least one honest server; later protocols may differ materially.

SRC-SUCHO

Saving Ukrainian Cultural Heritage Online

Publisher
SUCHO
Class
Professional or project methodology
Jurisdiction or scope
Cross-border cultural preservation
Retrieval
Not independently reopened in this release
Next review

Authority: Project documentation for distributed emergency preservation of cultural websites and digital heritage threatened by war.

Limit: One emergency-preservation project cannot establish a universal archive model; legal authority, consent, custody, and safety differ by collection.

SRC-SYRIAN-ARCHIVE

Syrian Archive — preservation and verification programme

Publisher
Mnemonic / Syrian Archive
Class
Professional or project methodology
Jurisdiction or scope
Syrian and international accountability work
Retrieval
Not independently reopened in this release
Next review

Authority: Project documentation for preserving, verifying, and organizing digital evidence at risk of loss.

Limit: Project material demonstrates one model and should not be generalized to all archives without review.

SRC-TAYLOR-FUNCTION-CREEP

The concept of function creep

Publisher
Law, Innovation and Technology
Class
Original peer-reviewed research
Jurisdiction or scope
Socio-legal research
Retrieval
Retrieval blocked or unavailable on 2026-09-02
Source date
Next review

Authority: Scholarly analysis of function creep in data-processing systems and institutional practice.

Limit: Direct page retrieval failed during this review; the public synthesis uses only a high-level concept until the full text is rechecked.

SRC-TAYLOR-HYPERNUDGING

Hypernudging — Big Data as a mode of regulation by design

Publisher
Information, Communication & Society
Class
Original peer-reviewed research
Jurisdiction or scope
Socio-legal research
Retrieval
Retrieval blocked or unavailable on 2026-09-02
Source date
Next review

Authority: Scholarly source associated with the concept of dynamic, data-driven choice architecture.

Limit: Direct page retrieval failed during this review; claims should be checked against the full article before precise reliance.

SRC-TOR-CENSORSHIP

Tor Project — censorship circumvention support

Publisher
The Tor Project
Class
Professional or project methodology
Jurisdiction or scope
Global anti-censorship technology
Retrieval
Not independently reopened in this release
Next review

Authority: First-party documentation about bridge and anti-censorship concepts for access under network blocking.

Limit: Availability and safety vary by jurisdiction, threat model, and current blocking techniques.

SRC-UAIX-COGNITIVE-LIBERTY-CHARTER

UAIX — Cognitive Liberty Charter Draft

Publisher
UAIX
Class
Professional or project methodology
Jurisdiction or scope
Worldwide conceptual framework
Retrieval
Not independently reopened in this release
Next review

Authority: First-party charter framing lawful inquiry, mental privacy, adult agency, persona integrity, and transparent boundaries.

Limit: A project charter is a normative source and does not itself establish law, empirical prevalence, or provider behavior.

SRC-UNESCO-AI-ETHICS

UNESCO Recommendation on the Ethics of Artificial Intelligence

Publisher
UNESCO
Class
Primary law or court record
Jurisdiction or scope
Global normative framework
Retrieval
Not independently reopened in this release
Next review

Authority: Intergovernmental recommendation addressing human rights, agency, privacy, transparency, and AI governance.

Limit: Recommendations require national implementation and do not prove compliance by any provider.

SRC-UNESCO-MEMORY-WORLD

UNESCO — Memory of the World Programme

Publisher
UNESCO
Class
Professional or project methodology
Jurisdiction or scope
Global documentary heritage
Retrieval
Not independently reopened in this release
Next review

Authority: International programme and principles supporting preservation and access to documentary heritage.

Limit: The programme does not resolve every conflict over authenticity, ownership, or access.

SRC-UNESCO-SAFETY-JOURNALISTS

UNESCO — Safety of Journalists

Publisher
UNESCO
Class
Official oversight record
Jurisdiction or scope
Worldwide journalism and human rights
Retrieval
Not independently reopened in this release
Next review

Authority: Intergovernmental programme and records concerning journalist safety, impunity, media freedom, and enabling conditions for reporting.

Limit: Programme material varies by publication and reporting period and should not be treated as a complete record of every attack.

SRC-UNGP-BUSINESS-HUMAN-RIGHTS

UN Guiding Principles on Business and Human Rights

Publisher
United Nations
Class
Primary law or court record
Jurisdiction or scope
Global business and human-rights framework
Retrieval
Not independently reopened in this release
Next review

Authority: International framework for state duties, corporate responsibility, due diligence, and remedy.

Limit: The principles are not themselves a universal enforcement mechanism.

SRC-US-HOUSE-OPERATION-CHOKE-POINT

U.S. House Oversight report — Operation Choke Point

Publisher
U.S. House Committee on Oversight and Government Reform
Class
Official oversight record
Jurisdiction or scope
United States
Retrieval
Not independently reopened in this release
Source date
Next review

Authority: Legislative-committee account of alleged informal pressure on financial institutions.

Limit: A committee report reflects the committee majority’s investigation and framing; underlying records and later developments should be reviewed for contested claims.

SRC-VDEM-DEMOCRACY-REPORTS

V-Dem Institute — Democracy Reports

Publisher
V-Dem Institute
Class
Civil-society or secondary analysis
Jurisdiction or scope
Global comparative research
Retrieval
Not independently reopened in this release
Next review

Authority: Cross-national research and indicators concerning democracy, expression, media, and autocratization.

Limit: Composite indicators depend on methodology, expert coding, uncertainty intervals, and report year.

SRC-WIKILEAKS-BANKING-BLOCKADE

WikiLeaks — Banking Blockade information package

Publisher
WikiLeaks
Class
Expert interview or commentary
Jurisdiction or scope
Cross-border payment access case study
Retrieval
Not independently reopened in this release
Next review

Authority: First-party account and supporting material concerning the payment blockade imposed on WikiLeaks.

Limit: The affected organization is not an independent adjudicator; claims require comparison with payment-provider, court, and regulator records.