Anthropic's Mythos AI Model Can Find High-Severity Vulnerabilities Across Major Systems

Key Points

• Anthropic's Mythos Preview can find vulnerabilities in every major OS and browser
• The model improves exploit development capabilities for both defenders and attackers
• IMF chief warns time is running out to address AI-driven cyber risks
• Cybersecurity stocks fell amid fears AI tools could disrupt traditional defenses

Full Details

Anthropic announced its Mythos Preview AI model this week, claiming it can find high-severity vulnerabilities across all major operating systems and web browsers. The model not only identifies security flaws but is also better at developing exploitation methods, which could empower both defenders and malicious actors. This development has prompted the IMF's managing director to warn that 'time is not our friend' regarding cyber risks, while Federal Reserve and Treasury officials held urgent meetings with Wall Street leaders. Security experts note that such models could reshape cybersecurity by giving developers tremendous scale in identifying vulnerabilities. The advancement has already triggered market reactions, with cybersecurity stocks falling as traders fear AI tools could upend traditional defenses.