The IETF aipref working group is defining a vocabulary for expressing preferences about how content may be used by AI systems, and how to attach those preferences to content, including through robots.txt and HTTP headers.
AI Preferences (aipref)
The IETF aipref working group is defining a vocabulary for expressing preferences about how content may be used by AI systems, and how to attach those preferences to content, including through robots.txt and HTTP headers.
Where it stands
As of 2026-10-13, the Agentic Standards Map records:
The vocabulary draft is at -08 (2026-09-13 in the datatracker history) and the attach draft was last seen at -05 (2026-08-19). A virtual interim meeting is announced for 2026-10-21. On 2026-10-13 the datatracker record gives revision 08 of the vocabulary draft (time 2026-09-14 06:18 UTC) and revision 05 of the attach draft (2026-08-19), no later revision; the interim announcement was not rechecked.
Why it matters to agents
The Working Group (WG) develops building blocks to express preferences about how content is collected and processed for AI model development, deployment and use, and the draft W3C Agentic Web charter lists it for coordination. For a seller, the vocabulary is the layer in which a publisher states what automated systems may do with its content. …
Sources
- https://datatracker.ietf.org/wg/aipref/about/
- https://datatracker.ietf.org/doc/draft-ietf-aipref-vocab
- https://datatracker.ietf.org/doc/draft-ietf-aipref-vocab/history/
- https://mailarchive.ietf.org/arch/msg/ai-control/H8Gycx8vy4BKierBvIhG8V2JYFQ/
- https://mailarchive.ietf.org/arch/msg/ai-control/A63f3A3k2a9GjhEtOy7J56aUorM
Status and relevance quoted, shortened, from the Agentic Standards Map by Francesco Marinoni Moretto, CC BY 4.0, https://frmoretto.github.io/agentic-standards-map/. The full entry, with every dated source, is in map.json under ietf-aipref.
Industry: cross-domain
Related standards
Standards this one builds on, supersedes, profiles or is commonly deployed beside.
