AI discovers thousands of zero-day vulnerabilities

Anthropic's Claude Mythos has demonstrated an alarming capability to autonomously discover thousands of zero-day vulnerabilities across major operating systems and browsers, significantly collapsing exploitation timelines. This advancement challenges traditional enterprise patching processes, which are often too slow to respond to the rapid pace of AI-driven exploits. Researchers found that AI models can now identify and potentially exploit vulnerabilities much faster than previously thought, with some exploits occurring within hours of disclosure, far outpacing the median five-day window for official listing and patching.