Anthropic Launches Project Glasswing with Apple and Google to Test AI Cybersecurity Using Claude Mythos Preview

Anthropic has initiated Project Glasswing, a collaborative effort involving Apple, Google, and over 45 additional organizations. The project will utilize the newly developed Claude Mythos Preview model to evaluate and enhance artificial intelligence cybersecurity capabilities. This initiative represents a significant partnership between major technology companies and numerous other entities focused on advancing security measures in AI systems. The collaboration aims to test cutting-edge cybersecurity approaches through the application of Anthropic's latest model. Project Glasswing brings together industry leaders and multiple organizations to address potential vulnerabilities in artificial intelligence technologies. The effort focuses on using the Claude Mythos Preview specifically for cybersecurity testing purposes. This marks a coordinated approach among competing AI labs and technology firms to strengthen digital security frameworks.