Anthropic scales Project Glasswing to 15+ countries

On Tuesday, artificial intelligence firm Anthropic announced it is expanding Project Glasswing, its joint industry initiative to find and fix critical software vulnerabilities using AI. The expansion will bring the program to about 150 new organizations across more than 15 countries. This scaling effort follows a period of significant financial momentum for the company. Anthropic recently filed confidentially for an initial public offering. This filing came after a massive $65 billion funding round that valued the company at nearly $1 trillion.

Project Glasswing relies on Claude Mythos, an AI model designed for vulnerability detection. In early April, Anthropic gave 50 initial partners, including the U.S. government, access to Claude Mythos Preview to scan their codebases for vulnerabilities and security flaws. The newly expanded cohort extends access to organizations managing critical infrastructure, including power, water, healthcare, communications, and hardware. The expansion covers several countries friendly to the U.S., including Australia, Canada, France, Germany, Italy, Switzerland, the Netherlands, Spain, Belgium, Sweden, India, Japan, New Zealand, and South Korea. Many of these organizations maintain codebases that other governments and entities rely upon.

The expanded list of partners includes U.S.-based identity and security management company Okta; South Korean companies Samsung, SK Hynix, and SK Telecom; the North Atlantic Treaty Organization (NATO); and the European Union Agency for Cybersecurity (ENISA). Anthropic emphasized the high stakes of securing these entities, warning that a successful attack on their codebase could be catastrophic. “For most partners, we estimate that a major attack could affect more than 100 million people, with important ramifications for both global and national security,” the company stated.

Anthropic is racing to establish these safeguards within Project Glasswing as competition in AI-driven cybersecurity intensifies. Rival OpenAI recently released its own cybersecurity-focused model, GPT-5.5-Cyber, which it has rolled out to a large group of partners for testing. Anthropic has stated it expects other AI companies to soon develop models as capable as Mythos Preview, making the rapid deployment of these security measures a priority for the firm.

Why it matters

Anthropic is positioning its AI as a critical layer for global infrastructure security, aiming to secure codebases that support millions of people while simultaneously signaling its massive scale via recent financial milestones.