Daily Tokens
November 16, 2026

AI · November 16, 2026 · High impact

Google reports the first fully automated cyberattack, run by an AI framework with no human involvement

Google's Threat Intelligence Group says it has for the first time caught attackers finding and exploiting security vulnerabilities with no human involved. The AI framework, called 'exploit_agent', was found on an exposed server. It automatically compared code before and after patches, built exploits, and ran them on its own.

It shows an end-to-end agentic vulnerability pipeline operating in the wild, which changes the threat model for patching and defense.

Go deeper

More from AI on Monday, November 16