Anthropic debuts powerful new AI model Mythos

Anthropic has launched a preview of its potent new AI model, Mythos, as part of its Project Glasswing cybersecurity initiative. This model, considered one of Anthropic's most powerful, is being deployed by over 40 partner organizations for defensive security tasks and to secure critical software. While not exclusively trained for cybersecurity, Mythos has reportedly identified thousands of zero-day vulnerabilities, some dating back decades. The model, integrated into Anthropic's Claude AI systems, possesses strong coding and reasoning skills, with partners including Amazon, Apple, Microsoft, and Google. The preview is not publicly available, but insights gained will benefit the broader tech industry.