Anthropic Unveils AI Cybersecurity Initiative with Major Tech Partners

Anthropic has launched Project Glasswing, a cybersecurity initiative partnering with major tech firms including Amazon, Microsoft, Apple, CrowdStrike, Palo Alto Networks, Google, and Nvidia.

The program grants selected organizations early access to Claude Mythos Preview, an unreleased AI model designed for defensive cybersecurity purposes.

Anthropic claims the model identified thousands of vulnerabilities across operating systems and web browsers. The startup plans to share findings with the industry and extend access to 40 critical infrastructure firms.

In addition, Anthropic pledged up to $100 million in usage credits and $4 million to open-source security initiatives. The company also said it is discussing the model’s capabilities with the U.S. government.

This move comes after recent reports that Claude Mythos posed both risks and advanced capabilities-prompting sharp declines in cybersecurity stocks like Palo Alto Networks and CrowdStrike.

The announcement coincided with heightened concerns about AI-powered cyberattacks at this year’s RSA Conference in San Francisco.