Claude Mythos AI Model Uncovers 271 Zero-Day Vulnerabilities in Firefox

Key Points:

• Anthropic's AI model, Claude Mythos Preview, identified 271 zero-day vulnerabilities in Mozilla Firefox, leading to the largest batch of security fixes in the browser's history with the Firefox 150 release.
• Since February 2026, Mozilla has collaborated with Anthropic to scan Firefox's codebase using frontier AI models, starting with Claude Opus 4.6 which found 22 vulnerabilities, including 14 high-severity ones, fixed in Firefox 148.
• Claude Mythos operates autonomously to find and exploit zero-day vulnerabilities across major operating systems and browsers, achieving high benchmark scores and successfully exploiting a significant portion of identified Firefox vulnerabilities.
• The AI-driven discovery marks a significant shift in cybersecurity defense by enabling faster, systematic, and more affordable vulnerability detection, potentially closing the historical gap favoring attackers.
• Beyond Firefox, Mythos uncovered long-standing vulnerabilities in critical infrastructure projects like OpenBSD, FFmpeg, and FreeBSD, highlighting AI's ability to reveal latent security risks overlooked for decades.