Svmuu News Anthropic's latest AI model, Claude Mythos, identified 271 security vulnerabilities in Mozilla's Firefox browser during testing. The related issues have recently been fixed, highlighting the potential of AI in the field of cybersecurity.
Mozilla stated that the model can efficiently analyze large-scale codebases and locate potential vulnerabilities, a process that previously heavily relied on scarce senior security researchers. In previous tests, another Claude model had discovered 22 critical vulnerabilities.
Nevertheless, Mozilla believes that the vulnerabilities discovered by AI are still within the capabilities of top human researchers, but its scalability may change the landscape of offense and defense, giving the defensive side a systematic advantage for the first time. However, the industry also warns that if such technology is misused, it could exacerbate the risk of automated attacks. (Decrypt)