Cloudflare starts a coalition to help websites recognize AI agents acting for users
On November 5 Cloudflare described the problems that personal AI agents create for websites and announced Trusted Agent Delegation, a group that will bring proposals to the IETF standards body. When an agent browses for someone, the site sees the agent instead of the person, which makes fraud checks, logins and account limits harder. Cloudflare also listed tools that site owners can already use to see and control agent traffic.
Read the original on Cloudflare Blog — Trusted Agent Delegation ↗
Why it matters
More shopping, booking and research is being done by AI agents on behalf of people. Without a shared way to know who an agent works for, sites must choose between blocking useful agents and accepting more abuse.
What changes
- For developers
- The proposals build on Web Bot Auth, where agents sign their requests, and on MOLE, an IETF draft for rate limiting without tracking individual users. Teams that build agents can join the work through the coalition's website.
- For businesses
- Cloudflare customers can already see personal-assistant traffic in Security Analytics and set access rules per agent. Account Abuse Protection, in early access, shows which agent made each login or sign-up.
Sources
- Cloudflare Blog — Trusted Agent Delegation ↗ blog.cloudflare.com Primary source
Summarized by AI from official sources on November 6, 2026. Spotted a mistake? Let us know. How it works