Cloudflare argues websites need new ways to identify and hold accountable AI agents

The post argues that hosted AI agents act as intermediaries, which removes a site's ability to identify, serve and hold accountable the user behind traffic and shifts costs onto the site as agentic traffic grows. It proposes an Intermediation Principle requiring identity, reputation and accountability mechanisms for both users and the agents and operators acting for them. It does not describe a specific vulnerability, CVE or patch, so no defensive action is required.

For security analysts · Written by AI from 1 source · Wrong?