How Anthropic’s Mythos has rewritten Firefox’s approach to cybersecurity

Mozilla says Anthropic’s Mythos AI helped uncover many high-severity Firefox vulnerabilities, including long-hidden sandbox and HTML parsing bugs.

The result was a surge to 423 bug fixes in April 2026, up from 31 in the same month a year earlier.

Mozilla says AI now helps researchers find issues and draft patches, but human engineers still review and implement the fixes.

The development highlights how agentic AI tools could strengthen software defense, while also creating new concerns if attackers use similar techniques.