Anthropic has launched Claude Mythos Preview, an advanced AI model capable of identifying and exploiting software vulnerabilities. Currently in preview status, it is not publicly available due to concerns about potential misuse. Instead, it is being utilized through Project Glasswing, a collaborative effort with tech giants including Amazon Web Services, Apple, Google, and Microsoft. This initiative aims to scan and strengthen critical software infrastructure across sectors like banking, healthcare, and logistics. Anthropic has pledged up to $100 million in usage credits and $4 million in donations to open-source security organizations. The model has already identified thousands of vulnerabilities, including significant flaws in major operating systems. As AI continues to evolve, Project Glasswing represents a proactive approach to cybersecurity, aiming to turn advanced AI capabilities into defensive tools against cyber threats. Anthropic plans to gradually expand access to the model while developing additional safety measures.