Anthropic's AI model finds bugs faster than Microsoft's ability to repair them.

In mid-May, dozens of Microsoft engineers and senior executives gathered at the company's headquarters in Redmond, Washington, to deliberate on the critical status of "Project Glasswing." The main topic of this meeting was the staggering speed of a new AI model named "Mythos" in discovering security vulnerabilities. The Mythos model, developed by the leading company Anthropic, identifies software vulnerabilities with unprecedented speed. Anthropic has provided this tool to a limited number of organizations so that these gaps can be closed before hackers affiliated with hostile governments use similar tools for espionage and sabotage. However, reports indicate that this model is finding weaknesses even faster than Microsoft’s technical teams can fix the bugs.