AI Model Discovers 22 Firefox Vulnerabilities in Two Weeks

Anthropic's Claude Opus 4.6 discovered 22 Firefox security vulnerabilities in two weeks, including 14 high-severity bugs representing nearly 20% of all high-severity Firefox bugs patched in 2025. Mozilla validated the findings and shipped fixes in Firefox 148. Beyond discovery, Claude also wrote working exploits for two bugs after roughly $4,000 in API credits and hundreds of attempts. The AI found new classes of logic errors that traditional fuzzing never caught. Anthropic notes that frontier models are now world-class vulnerability researchers, but warns the gap between discovery and exploitation capabilities is narrowing fast, signaling an accelerating arms race in cybersecurity. Mozilla is now integrating AI-assisted analysis into its security workflows.