Cloudflare launches adaptive application security framework to counter AI-driven threats

Cloudflare has announced a new adaptive application security framework designed to address the evolving threat landscape driven by AI agents. The initiative connects four key areas: discovery, governance, runtime protection, and response, aiming to move beyond siloed security tools. Key launches include Application Profiles, which automatically learn normal application behavior to enforce a positive security model, and LLM-hardened WAF rules, where Cloudflare partnered with major model providers to use frontier models for adversarial validation and bypass detection. Additionally, access to Cloudforce One’s Threat Events Platform is now free for all Cloudflare accounts, enabling customers to block requests from compromised infrastructure. To manage the rise of autonomous agents, Cloudflare introduced Precise Session Analysis and Adaptive Intelligence to distinguish between legitimate automation and malicious activity. The company is also developing automated security operations capabilities to correlate alerts and recommend mitigations, though these are currently in development with the Managed Defense team.