Anthropic’s AI Finds Software Flaws Faster Than Microsoft Can Fix Them

In mid-May, Microsoft engineers gathered to confront software weaknesses uncovered at an unprecedented pace by Anthropic’s new AI model, Mythos, as part of Project Glasswing. The AI firm provided select organizations access to the tool to identify vulnerabilities before hackers or adversarial governments could exploit them. During the meeting, one engineer asked whether Mythos lived up to the hype Anthropic had promoted.