Anthropic launched Project Glasswing, giving a coalition of major technology and cybersecurity firms access to a preview version of Claude Mythos. The initiative is designed to strengthen defenses by using the model to scan systems and open-source code for vulnerabilities ahead of broader deployment. The effort includes organizations such as Amazon Web Services, Apple, Google, Microsoft, JPMorganChase, CrowdStrike, and NVIDIA. Anthropic said it does not plan to make the preview generally available, but expects safer scaling once safeguards are in place. Campus and higher education IT leaders may take note because the project’s approach—using frontier models for proactive vulnerability discovery—signals where institutional security teams are likely headed as generative AI capabilities expand.