Back to AIBriefs
EventCybersecurityPolicy

Anthropic Mythos model finds vulnerabilities in US govt systems

The Mythos model identified vulnerabilities within hours during a testing exercise with US intelligence agencies, an official told the AP. The exercise was part of Project Glasswing, which aims to secure critical software from potential AI fallout. However, the model did not necessarily exploit the discovered flaws.

Jun 24, 3:29 AM