Reported Platform & EcosystemAccess ControlsStandards & Protocols Cloudflare Blog ·

Cloudflare Unveils AI-Era Security Framework and Expands Bot Governance Tools

Open source ↗

Summary

Cloudflare introduced a four-stage application security framework designed to address autonomous AI-driven cyber attacks and govern agentic web traffic. As part of the release, the company announced Application Profiles to enforce positive security models automatically and made its Cloudforce One threat events intelligence accessible to all accounts. The architecture also expands Botbase to establish verified identities and behavioral trust tiers for automated bots and AI agents.

Insight

Edge enforcement is shifting from binary bot-blocking scores toward multi-dimensional identity and session trust models that distinguish beneficial AI agents, standard search crawlers, and autonomous adversarial exploits.

Implication

Publishers and site operators using Cloudflare can apply granular policies to separate search indexing, training scraping, and transactional user agents, while custom AI agents must adopt verified registration workflows to maintain access.

Why it matters

As autonomous AI agents blur the line between human browsing, routine web crawling, and automated exploits, proxy providers are restructuring the foundational access rules that govern automated machine interaction across the web.

Evidence