Researchers from Calif, a cybersecurity firm based in Palo Alto, California, claim to have breached Apple's macOS by developing a privilege escalation exploit with the assistance of Anthropic's Claude Mythos Preview. This exploit potentially allows unauthorized access to restricted sections of Mac computers, enabling attackers to take control. The researchers utilized Mythos to identify vulnerabilities quickly, demonstrating the capability of advanced AI systems to uncover previously unknown security flaws. Apple has responded to these findings with seriousness, stating that security is their top priority and engaging with the researchers at Apple Park in Cupertino to discuss the implications of what they describe as the 'first public macOS kernel memory corruption exploit on M5 silicon.' The researchers plan to release detailed technical information after Apple addresses the vulnerabilities. Anthropic's Mythos is part of Project Glasswing, which aims to enhance cybersecurity measures across various organizations, including major tech firms like Amazon Web Services and Microsoft. This initiative reflects a growing trend where AI is leveraged not only for identifying vulnerabilities but also for strengthening overall cybersecurity frameworks.