Anthropic AI Finds Critical Software Flaws Worldwide
Anthropic’s Claude Mythos AI found more than 10,000 high- and critical-severity vulnerabilities in some of the world’s most widely used software, just one month after Project Glasswing launched. Project Glasswing is Anthropic’s AI-powered cybersecurity initiative. Through it, roughly 50 partner organizations gained access to Claude Mythos Preview, a frontier AI model built to detect vulnerabilities […]



