AI Business
Jun 2, 2026
Anthropic expands AI vulnerability detection initiative to 150 organizations in over 15 countries
Jun 2, 2026
AI Summary
Anthropic is broadening its Project Glasswing initiative, which uses AI to identify software vulnerabilities, to approximately 150 organizations across more than 15 countries. This expansion includes critical sectors such as power, healthcare, and communications, aiming to enhance security for systems that impact millions of people.
- Anthropic is extending its Project Glasswing initiative to around 150 new organizations in over 15 countries. This initiative focuses on identifying and fixing critical software vulnerabilities using AI technology.
- The expansion follows Anthropic's confidential filing for an initial public offering after a significant funding round valuing the company at nearly $1 trillion.
- The AI model at the center of this initiative, Claude Mythos, is designed to detect thousands of zero-day vulnerabilities over several weeks. Initial access was granted to 50 partners, including the U.S. government, in April.
- The newly included organizations span various critical industries, such as power, water, healthcare, and communications, which were previously underrepresented in the initiative.
- Anthropic emphasized that a successful attack on the codebases of these partners could have catastrophic consequences, potentially affecting over 100 million people.
- The expanded list includes organizations from countries allied with the U.S., such as Australia, Canada, and several European nations, as well as companies like Okta, Samsung, and SK Telecom.
- Anthropic is racing to establish safeguards within Project Glasswing, anticipating that other AI companies will soon develop similarly capable models, following the release of OpenAI's cybersecurity-focused model GPT-5.5-Cyber.
anthropiccybersecuritycritical infrastructuremythosproject glasswing