💨 Abstract

Anthropic has released a preview of its new AI model, Mythos, for cybersecurity work through Project Glasswing. This model, part of Anthropic's frontier series, is designed for complex tasks like coding and reasoning. Over 40 partners, including Amazon and Microsoft, will use Mythos to scan software for vulnerabilities. The model has already identified thousands of critical issues. Despite a recent data leak, Anthropic assures that Mythos will be used defensively.

Courtesy: Lucas Ropek