Mozilla said it had identified and fixed 271 security vulnerabilities in Firefox 150, thanks to early access to Mythos, an AI model developed by Anthropic with advanced capabilities for analyzing code and detecting vulnerabilities.

According to Bobby Holley, Firefox's chief technology officer, the results represent an important shift in cybersecurity, giving defenders tools capable of competing effectively with attackers.

Although the severity of the vulnerabilities was not disclosed, Holley said Mythos had clearly outperformed another Anthropic model, which identified only 22 vulnerabilities in an earlier Firefox release.

He explained that many of these flaws could have been found through fuzz testing or by professional security researchers. Mythos, however, saved months of human work, reducing costs and improving the efficiency of threat detection.

Holley said AI now plays a central role in strengthening cybersecurity. These tools can compete with the world's leading researchers, he added, and this type of analysis will become an essential part of software development in the future.