Anthropic scales Claude Mythos to critical infrastructure in 15+ countries

Anthropic is significantly expanding Project Glasswing, an initiative that utilizes AI to identify and resolve critical software vulnerabilities. The program, built around Anthropic’s powerful Claude Mythos model, is now being rolled out to approximately 150 new organizations across more than 15 countries. This expansion addresses sectors that were not fully represented in the initial pilot, including power, water, healthcare, communications, and hardware industries. Many of these new partners maintain codebases crucial to other organizations and governments.

The Claude Mythos model, described by Anthropic as its most potent to date, is capable of detecting thousands of zero-day vulnerabilities. It was initially provided to 50 partners, including the U.S. government, for scanning their codebases. The current expansion includes organizations in allied nations such as Australia, Canada, France, Germany, Italy, Japan, and South Korea, as well as critical entities like NATO and the EU’s cyber security agency ENISA.

Anthropic emphasizes the critical nature of this expansion, stating that a successful attack on any partner’s codebase could be catastrophic. The company estimates that a major security breach could impact over 100 million people, with serious implications for both national and global security. This push to establish safeguards is timely, as rival AI companies are also developing similar cybersecurity models.

The expansion follows a substantial funding round for Anthropic, reportedly valuing the company at nearly $1 trillion. This comes as the company is said to be confidentially preparing for an initial public offering.