Anthropic Expands AI Cybersecurity Program to 150 More Partners While Warning Mythos Could Speed Up Attacks on Critical Infrastructure

Anthropic on Tuesday said it is expanding Project Glasswing to an additional 150 partners in more than 15 countries. The new partners will gain access to the Mythos model once they meet security requirements. The expansion covers industries that were not well represented in the initial launch, including power, water, healthcare, communications and hardware.

Anthropic rolled out initial testing of Mythos to 50 partners in April. Since the initial launch, Project Glasswing partners have unveiled more than 10,000 high or critical security flaws. The company estimates that a major cyberattack could impact more than 100 million people.

The announcement came a day after Anthropic said it would start offering access to the European Union. On Monday the company also confidentially filed its initial public offering prospectus with the Securities and Exchange Commission, beating rival OpenAI to the milestone. Major Project Glasswing partners include Apple, Nvidia, Microsoft, CrowdStrike and Palo Alto Networks.

Anthropic did not disclose the new companies joining the coalition. The White House held numerous meetings with Big Tech leaders and financial institutions over ways to safely deploy advanced AI models. Many experts warned that advanced cybersecurity capabilities existed beforehand but Mythos could allow hackers to expose vulnerabilities more quickly.