W3C paper proposes standards work for AI agents on the web
Start here
W3C's Strategy Team published a paper arguing that AI agents have moved from reading the web to interacting with it, so long-standing assumptions about crawling, traffic and consent no longer hold. It proposes a declaration of web values for the AI era and next steps including chartering an Agentic Web Working Group and pairing agent requirements with user requirements across W3C groups. The paper cites vendor data that in 2026 more than half of traffic on Cloudflare's network was not human-initiated, and that daily AI agent requests grew over 1,700% in the year to September 2026. It also warns that prompt injection, delegated consent and publisher sustainability remain unresolved. The Strategy Team says the paper contains early recommendations it will present at the web and AI IG session on 29 October during TPAC 2026 and submit to each W3C group for discussion, with feedback also accepted by email and in the w3c/webai-roadmap repository.
Generated from reports · updated 39 min ago
Report timeline
Follow the reports to see the event from different sides.
- Hacker News · AIW3C paper examines how AI agents are changing the web and proposes standards work
W3C's Strategy Team published a paper arguing that AI agents have moved from reading the web to interacting with it, and that the web's long-held assumptions about crawling, traffic and consent no longer hold. It cites vendor data showing that in 2026 more than half of traffic on Cloudflare's network was not human-initiated and that daily AI agent requests grew over 1,700% in the year to September 2026, while warning that prompt injection, delegated consent and publisher sustainability remain unresolved. The paper proposes a declaration of web values for the AI era and next steps including chartering an Agentic Web Working Group and pairing agent requirements with user requirements across W3C groups.
Heat for this event
Not enough continuous observations yet to draw a trend.