Anthropic Unveils Restricted AI Cyber Model in Unprecedented Industry Alliance

Anthropic launched Project Glasswing, a collaboration with major tech and security firms to use its Claude Mythos Preview AI model to identify software vulnerabilities at scale; the company reports thousands of serious flaws found across operating systems, browsers, and core components, has coordinated disclosures and patches, and is withholding public release of the model due to misuse risks while urging industry-wide defensive cooperation.