Anthropic Expands Claude Mythos to 150 Critical Infrastructures

Anthropic recently announced a major expansion of its Project Glasswing initiative, which aims to collaborate with the industry to identify and rectify critical software vulnerabilities. This expansion involves approximately 150 new organizations across more than 15 countries, according to the announcement made by the company on Tuesday.

This announcement comes shortly after Anthropic indicated that it had filed a confidential application for an initial public offering, following a funding round of $65 billion, bringing its valuation to nearly $1 trillion. This development underscores the growing importance of the company in the field of cybersecurity.

The Claude Mythos model is at the heart of Project Glasswing. Anthropic describes it as its most powerful model to date, capable of identifying thousands of zero-day vulnerabilities in just a few weeks. Earlier this year, Anthropic provided access to 50 initial partners, including the U.S. government, to Claude Mythos Preview to scan their codebases for vulnerabilities and security flaws.

Today, the expanded list of organizations with access to Mythos spans sectors such as energy, water, health, communications, and hardware. These sectors were not "well represented" in Anthropic's initial group, the company clarified. Many of those who will now have access are businesses or NGOs that maintain codebases relied upon by other organizations and governments, Anthropic noted in its blog post.

"What every partner has in common is that a successful attack on their codebase could be catastrophic," the company stated. "For most partners, we estimate that a major attack could affect over 100 million people, with significant implications for global and national security."

The expanded group includes organizations in U.S. ally countries, including Australia, Canada, France, Germany, Italy, Switzerland, the Netherlands, Spain, Belgium, Sweden, India, Japan, New Zealand, and South Korea, according to the Financial Times, citing a source close to the matter.

The Financial Times also reported several organizations that have gained access to Mythos, including the U.S.-based identity and security management tool Okta; South Korean companies Samsung, SK Hynix, and SK Telecom; NATO, the U.S.-led military alliance based in Brussels; and the EU cybersecurity agency ENISA.

Anthropic stated that it expects other AI companies to soon develop models as capable as Mythos Preview, which is why the company is working to establish protective measures within Project Glasswing. Since the release of Mythos, competitor OpenAI has launched its own cybersecurity-focused model, GPT-5.5-Cyber, which it has deployed to a broad group of partners for testing.