Mozilla says “defenders have a chance to win” after testing Claude Mythos in Firefox

Mozilla has stated that with the use of the artificial intelligence (IA) “the defenders have a chance to win” against cybercriminals, and has done so after identifying and correcting more than 270 vulnerabilities in Firefox with the help of Claude Mythos, Anthropic’s security model.

The latest Firefox browser update (v150), available this week, includes fixes for 271 vulnerabilities that have been identified during a evaluation with the preliminary version of Claude Mythosa model created specifically to detect high severity failures in major operating systems and web browsers.

Zero-day vulnerabilities “have contact days” according to Mozilla, which sees artificial intelligence as the definitive tool to protect critical software exposed to the Internet, which until now depended on teams dedicated to security.

With AI, “defenders finally have the opportunity to win, decisively,” the technology company says in a statement. And he says this based on his experience with Anthropic’s latest models for discovering vulnerabilities.

Apart from Claude Mythos, Mozilla has also used Claude Opus 4.6 to overhaul its browser. In March, it announced the identification of 112 failures of different levels of severity, including 14 high severity.

“The gap between machine-detectable and human-detectable bugs favors the attacker, who can spend many months of costly human effort finding a single bug. Closing this gap reduces the attacker’s long-term advantage by making all discoveries cheaper.“Mozilla adds.