Anthropic is adding 150 new organizations to Project Glasswing, its program to find and fix security vulnerabilities in critical software.
The expansion, announced June 2, grows a program that began in early April with about 50 partners using Claude Mythos Preview, a frontier AI model Anthropic has withheld from public release over cybersecurity concerns. Those initial partners have surfaced more than 10,000 high- or critical-severity flaws while scanning their codebases.
The new cohort spans more than 15 countries and adds industries underrepresented from the first group, including healthcare, power, communications and hardware. Many new partners are vendors and open-source maintainers whose code other organizations, including governments, depend on. The company estimates a major attack on most partners could affect more than 100 million people.
A broader expansion is on the horizon.
“We’re working as quickly as we can to safely release Mythos-level capabilities in general access,” Anthropic’s press release noted. “To do so, we’ll need highly robust safeguards that prevent the model’s cyber capabilities from being misused – safeguards that we (and, to our knowledge, all other AI developers) have yet to develop. Because cybersecurity has both helpful and destructive uses, making safeguards that are both strong and precise enough is a major challenge.”
Anthropic also plans to scale up its Cyber Verification Program, granting additional capabilities to organizations for cyberdefenses. The company is in discussions with third parties to scale up the review and patching capabilities in open-source software, as companies with initial Claude Mythos Preview access have begun using the model to write patches and automate detection and response.
“In the future, frontier model releases will become increasingly high-stakes. Capabilities will continue to improve across all domains, including many that — like cybersecurity — can empower attackers and defenders alike,” the release notes. “This will not be the last time we need to confront a challenge like this one. But Project Glasswing has taught us a great deal about how to respond when models cross important capability thresholds. If we’re successful, we hope to enable a permanent advantage for defenders.”
The expansion deepens Anthropic’s healthcare push. The company launched the HIPAA-compliant Claude for Healthcare on Jan. 11, promising faster prior authorizations, better care coordination and quicker regulatory filings, with early adopters including Phoenix-based Banner Health, Chicago-based CommonSpirit Health and Palo Alto, Calif.-based Stanford Health Care.
Anthropic raised $65 billion at a $965 billion valuation May 28, with run-rate revenue crossing $47 billion, underscoring the scale behind its move into regulated sectors. On June 1, Anthropic confidentially submitted a draft Form S-1 registration statement to the SEC, a step toward a potential initial public offering that the company said would depend on market conditions.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
