Anthropic Deploys Claude Mythos AI to Guard Global Critical Infrastructure
Anthropic expands its Project Glasswing initiative to over 150 organizations in 15 countries, using its Claude Mythos AI model to hunt zero-day vulnerabilities in essential systems like power, water, and healthcare networks.
Anthropic expands Project Glasswing, its AI-driven initiative to find and fix critical software vulnerabilities, to approximately 150 new organizations across more than 15 countries. This massive expansion comes just one day after Anthropic confidentially files for an initial public offering following a $65 billion funding round that values the company at nearly $1 trillion. The Claude Mythos AI model powers this entire effort, serving as Anthropic's most powerful tool yet for identifying thousands of zero-day vulnerabilities over extended periods.
The newly added partners represent crucial sectors that previously lack representation in the program, including power, water, healthcare, communications, and hardware. Anthropic emphasizes that every selected organization maintains codebases where a successful cyberattack affects over 100 million people and causes severe global and national security ramifications. The expanded global roster spans U.S.-allied nations such as Canada, France, Germany, Japan, and South Korea, featuring major entities like Okta, Samsung, SK Hynix, NATO, and the EU's cybersecurity agency ENISA.
Anthropic acknowledges that rival AI companies will soon develop models matching Mythos Preview's capabilities, prompting a race to establish robust safeguards through Project Glasswing. The competitive landscape already shifts as OpenAI releases its own cybersecurity-focused model, GPT-5.5-Cyber, which rolls out to a large group of testing partners. By securing critical infrastructure first, Anthropic aims to set the global standard for AI-assisted cybersecurity before the market fully crowds with competing technologies.