US official: Anthropic broke into some of govt's most sensitive, secure systems

A powerful AI model, Mythos, has successfully identified vulnerabilities in sensitive US government computer systems within hours during a recent test. This revelation, part of Project Glasswing, has amplified concerns about advanced AI's capabilities and potential risks. Despite this cooperation, tensions persist between AI firm Anthropic and the Trump administration over access restrictions to its advanced models, with experts urging a reconsideration of these limitations.

By TOI TECH DESK