
Adapting to AI penetration testing orchestrators: The Modern Security Model
Cloudflare says a burst of AI-driven penetration-testing frameworks is now scanning its customers at a scale defenders are struggling to match. The open-source agentic tool ARTEX was taken down by its creator on Oct. 8 after malicious use, but in the month since more than 12,000 related repositories have appeared, and in just three days more than 50 framework families tried to scan Cloudflare customers worldwide from thousands of IP addresses. The company argues security teams need a model built for this volumetric, autonomously guided probing.
Summary written by Readthrough from the original article. The full story, with photos and any updates, is at Cloudflare Blog.



