On April 7 (local time), Anthropic, in partnership with 11 leading tech companies, unveiled Project Glasswing and introduced its state-of-the-art cybersecurity AI model, Claude Mythos Preview. This collaborative initiative involves a consortium of 12 industry titans, including Amazon Web Services, Anthropic, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorgan Chase, the Linux Foundation, Microsoft, NVIDIA, and Palo Alto Networks. Together, they aim to tackle the pressing cybersecurity challenges that have emerged in the AI-driven landscape.

The Claude Mythos Preview model stands out for its exceptional autonomous vulnerability detection capabilities. It has autonomously pinpointed thousands of high-risk zero-day vulnerabilities across all major operating systems and web browsers. Demonstrating its commitment to the project, Anthropic has pledged up to $100 million in model usage credits. Additionally, the company has donated $4 million to the Linux Foundation and the Apache Software Foundation to bolster support for open-source software maintainers.

At present, the model has been made accessible to over 40 organizations responsible for maintaining critical infrastructure. However, given its potential for dual-use—where it could be exploited for both beneficial and malicious purposes—there are currently no plans to release it to the general public.