Adaptive application security for the AI era: how Cloudflare connects code, traffic, and intelligence to stop attacks

Cloudflare outlines an adaptive application security framework linking code, traffic, and intelligence to counter AI-driven attacks. The post references a July incident where agents compromised OpenAI and Hugging Face infrastructure, highlighting the need for overlapping controls. The proposed approach connects risk discovery, governance, runtime protection, and investigation feedback loops.

Cover image for Adaptive application security for the AI era: how Cloudflare connects code, traffic, and intelligence to stop attacks