Anthropic’s Project Glasswing Enlists Nvidia to Hunt Thousands of Zero-Day Bugs
Anthropic’s Project Glasswing will let Nvidia join Microsoft and Amazon in testing Claude Mythos Preview, an AI model that has identified thousands of zero-day vulnerabilities in operating systems and web browsers. Anthropic committed $100 million in usage credits and $4 million in donations to open-source security groups under the initiative.
1. Project Glasswing Launch and Partners
Anthropic introduced Project Glasswing, granting select organizations including Nvidia, Amazon, Microsoft, Apple, CrowdStrike, Palo Alto Networks and Google access to its unreleased Claude Mythos Preview model for defensive cybersecurity testing.
2. Model’s Vulnerability Discoveries
Claude Mythos Preview has identified thousands of critical zero-day vulnerabilities across major operating systems, web browsers and other software, uncovering bugs some up to 27 years old.
3. Financial Support and Access Expansion
Under Project Glasswing, Anthropic pledged $100 million in usage credits and $4 million in donations to open-source security groups, and extended model access to about 40 additional critical software infrastructure organizations.
4. Regulatory Engagement and Long-Term Vision
Anthropic is in discussions with the U.S. government regarding the model’s offensive and defensive capabilities and aims to enable safe, large-scale deployment of Mythos-class AI models.