TechCrunch reported on Monday that Anthropic will expand Project Glasswing, an initiative to help organisations use its AI models to find vulnerabilities and prepare response measures, to 150 organisations in more than 15 countries.

In early April, Anthropic provided 50 initial partners, including the U.S. government, with access to Claude Mythos Preview.

With the move, new participants in Project Glasswing include companies in the power, water, healthcare, telecommunications and hardware sectors that were not heavily represented in the initial group.

Anthropic said partners share a common factor in that an attack on their codebase could have catastrophic consequences. It said a major attack could affect more than 100 million people.

The Financial Times reported that new participants include identity and security management tools company Okta, South Korean companies including Samsung Electronics, SK Hynix and SK Telecom, as well as NATO and the EU cyber security agency ENISA. It said participating countries added South Korea, Australia, Canada, France, Germany, Italy, Switzerland, the Netherlands, Spain, Belgium, Sweden, India, Japan and New Zealand.

TechCrunch reported that Anthropic is moving quickly to build safeguards through Project Glasswing because it expects other AI companies to develop models on par with Mythos Preview soon.