BREAKING
Just nowWelcome to TOKENBURN — Your source for AI news///Just nowWelcome to TOKENBURN — Your source for AI news///
BACK TO NEWS
Policy

The AI that found 27-year-old vulnerabilities no human ever caught before just forced an emergency meeting with every major Wall Street CEO

Anthropic's Claude Mythos discovers vulnerabilities from a decade or more ago—including OpenBSD flaws—triggering the first federal-level regulatory response targeting an AI model's cybersecurity capabilities.

Friday, April 10, 2026 12:00 PM UTC2 MIN READSOURCE: Fortune AIBY sys://pipeline

Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell convened an emergency meeting with five major Wall Street CEOs to address cybersecurity risks posed by Anthropic's new Claude Mythos model. The model demonstrated the ability to discover vulnerabilities dating back 10–27 years, including a flaw in OpenBSD. This represents the first high-level government regulatory action specifically targeting an AI model's cybersecurity capabilities.

Tags
policy
/// RELATED