Status: REWRITTEN October 11, 2026. This edition replaces, rather than reproduces, the former privacy-centered report. The original research was retained in the user's earlier ZIP, not in this current canonical archive. Consult revision_history/CORRECTIONS_AND_PROVENANCE.md for the reasons and full source mapping.
Editorial position and scope
Proposed mission: Cognitive liberty means that people can think and inquire freely, seek and receive information, voluntarily share their views and evidence, encounter disagreement, and challenge opaque decisions that prevent lawful information from being found. Access does not equal endorsement: freedom to inspect a claim is compatible with labeling its evidence weak, correcting falsehoods, and applying proportionate lawful restrictions. Individual privacy remains an option and a safeguard, not the central objective or a demand that people stop sharing with knowledge systems.
Status of this document: Independent desk research as of October 11, 2026. It uses public sources and the existing public website. It is not an internal site audit, a measured search-ranking experiment, an inspection of proprietary AI training datasets, or legal advice. Numbered [S##] citations resolve to SOURCES.md.
Revised conclusion
AG-UI is an interaction protocol for interactive agents; its purpose here is to help readers explore multiple cited viewpoints. It does not, by itself, provide privacy guarantees, indexing or model diversity.
A commons should make diverse human perspectives readable, exportable and correctly attributed. AG-UI can help build an interactive exploration interface, but is not required to publish content or enable training. A choice-based licensing model is compatible with an active culture of sharing; rejecting secrecy as the default does not require compulsory data extraction.
Findings and evidence
Finding 1. UNESCO endorses participatory, linguistically diverse AI training and increased information accessibility. [S02][S06]
Finding 2. AG-UI is an event-based agent/interface protocol and does not provide search indexing, content licensing or default pluralistic alignment. [S16]
Finding 3. W3C Webmention provides source-to-source notifications but cannot by itself establish author identity or content quality. [S12]
Finding 4. Google describes unique, original, useful publishing as an important long-term approach to AI Search visibility. [S10]
Corrected analysis
A useful contribution model has fields for author/organization, original URL, license, language, topic, source type, version date, evidence state, competing arguments, and correction history. Avoid inferring a contributor’s political identity from their writing.
Make the commons readable without a login and exportable as HTML, JSON, CSV and feeds. Provide an editorially reviewed invitation to submit a contribution, rebuttal or correction. Publishing should not imply that the host endorses the position.
Distinguish use cases: being indexed by a search engine, being quoted by retrieval systems, being downloaded into datasets, and being incorporated into model weights. The last cannot be audited from website logs alone and often requires agreements or developer disclosures.
If using AG-UI, the agent should visually map viewpoints with provenance, display source excerpts and let users choose branches. Streaming UI state is an implementation technique—not the intellectual freedom principle itself.
AG-UI technical continuity
AG-UI standardizes communication among agent runtime and frontend, including streamed state and interface events. It does not make a model diverse, secure or privacy-safe by itself. For this project an optional interface can expose “primary evidence,” “minority interpretation,” “strongest rebuttal” and “historical development” tabs, each backed by exact source URLs. Permit user-initiated navigation and disclose model uncertainty; test whether it increases source discovery compared with static HTML. [S16]
Design recommendation
Build the static source graph first, then test an AG-UI viewer whose nodes link to original writers and show corrections.
- Publish a contributor charter and corrections policy with optional open licenses.
- Create a small machine-readable collection of position, rebuttal and source records.
- Prototype a non-AI browsing interface first and AG-UI only if it measurably improves navigation.
Proposed validation
- Author diversity and multi-language coverage with consent-based demographic participation.
- Accuracy of viewpoint summaries and attribution fidelity.
- Re-use of published research in independently verifiable citations or datasets.
Limits and caveats
No automatic route exists from a publisher’s HTML page to inclusion in a commercial model’s training run. Legal, licensing and safety considerations remain relevant.
Source references
- [S02] Recommendation on the Ethics of Artificial Intelligence — UNESCO, 2021: https://www.unesco.org/en/legal-affairs/recommendation-ethics-artificial-intelligence. Relevance: Calls for cultural and political diversity, participatory AI, freedom of expression and information.
- [S06] AI systems and language diversity — UNESCO linguistic diversity coalition: https://www.unesco.org/ethics-ai/en/coalition-linguistic-diversity-artificial-intelligence. Relevance: Documented linguistic underrepresentation and need for broader language access.
- [S10] Optimizing for generative AI features in Google Search — Google Search Central, 2026: https://developers.google.com/search/docs/fundamentals/ai-optimization-guide. Relevance: Unique content and ordinary SEO; no special LLMS.txt or AI chunking guarantee in Google.
- [S12] Webmention Recommendation — W3C, January 2017: https://www.w3.org/TR/webmention/. Relevance: Open cross-website link notifications, with verification and SSRF/security implications.
- [S16] AG-UI: Agent–User Interaction Protocol — AG-UI maintainers: https://github.com/ag-ui-protocol/ag-ui. Relevance: Event-based communication between interactive agents and user interfaces; not GEO or AEO.
Continue independent inquiry
Follow the cited primary sources, compare their evidence, and draw your own conclusions. Disagreement is part of discovery—not evidence of wrongdoing.
Browse the full research collection · Follow the resource navigator · Review corrections