Anthropic is scaling up Project Glasswing with roughly 150 new partner organizations across more than 15 countries. The project uses the Claude Mythos Preview model to scan critical software for security flaws. The roughly 50 partners already on board have found more than 10,000 serious vulnerabilities so far, Anthropic says.
New partners come from sectors like energy, water, healthcare, and communications. Anthropic estimates a successful attack on any of their systems could affect over 100 million people. The EU's cybersecurity agency, ENISA, is also reportedly set to get access, though the terms are apparently still being negotiated.
Anthropic warns that rival AI companies could ship comparable models within 6 to 12 months, potentially without safeguards against misuse. Over the coming weeks, the company plans to make Mythos capabilities broadly available once it's built enough protections.
This could turn into a lucrative business for Anthropic. With Claude Security, the company already sells a commercial product built on Claude Opus 4.8 that scans codebases and suggests patches. In effect, it's selling protection against attacks from Mythos-class models, the very threat Anthropic itself is spotlighting through Project Glasswing.
