Cybersecurity Policy Report, Anthropic Expands AI-Powered Cyber Defense Collaborative, (Jun 2, 2026)
Anthropic has added 150 organizations to Project Glasswing, a cyber defense effort that uses Anthropic’s Claude Mythos2 AI (artificial intelligence) model to find and fix cybersecurity vulnerabilities, the company announced today.
Anthropic launched Project Glasswing in April after deciding that Claude Mythos2 was too dangerous for public release. Instead, it provided Mythos2 to a select group of organizations that were capable of using the AI model for cyber defense (CPR, April 8). Since then, 50 Project Glasswing participants have identified “more than 10,000 high- or critical-severity security flaws,” Anthropic noted today in a news release.
The Project Glasswing expansion will enlarge the group from about 50 participants to around 200. The new participants “are based in more than 15 countries, and most provide critical infrastructure to many more,” Anthropic said.
“The group covers several industries that weren’t well represented in our initial cohort, such as power, water, healthcare, communications, and hardware. And many of the new partners are vendors—companies or nonprofits that maintain codebases that are relied upon by lots of other organizations around the world, including governments,” it said.
“This expansion is the next step toward our long-term goals: for AI to make all software more secure, and for us to help the industry adjust to how AI could change many of the core assumptions of cybersecurity,” Anthropic said.
Within six to 12 months, AI models with the same hacking capabilities as Claude Mythos2 could be released by other developers “without safeguards that prevent misuse,” the company warned.
“In that world, cyberattacks could occur much more often, and in much more unpredictable forms. It’s imperative that cyberdefenders adapt to maintain pace,” it added. “Cheap, fast AI models with powerful cyber capabilities are around the corner. We want Project Glasswing to spur institutions toward operating norms that reflect this reality.”
News: InternationalLegislation DataSecurity AINews