Anthropic announced on Tuesday that it is expanding its Project Glasswing initiative to include 150 additional partner organizations across more than 15 countries. These new participants will test Mythos, a specialized AI model designed to identify software vulnerabilities. This move significantly broadens the program's reach beyond its initial launch, now incorporating critical sectors such as power, water, healthcare, communications, and hardware. All new partners must satisfy strict security requirements before gaining access to the model. The expansion represents a strategic step toward Anthropic's long-term vision of making all software more secure. The company aims to assist the industry in adapting to how AI might fundamentally shift core assumptions in cybersecurity. This development follows closely on the heels of Anthropic's decision to make Mythos accessible to the European Union. Additionally, earlier in the week, the company confidentially filed its IPO prospectus with the U.S. Securities and Exchange Commission, positioning itself ahead of rival OpenAI and setting the stage for a major public market offering. The program initially began testing with 50 partners in April. At that time, the rollout faced scrutiny as experts warned that such advanced AI capabilities could already exist and might be exploited by malicious actors to expose vulnerabilities more rapidly than traditional methods could. These concerns prompted numerous meetings between the White House, major technology firms, and financial institutions to discuss safe deployment protocols for advanced AI models in the cybersecurity sector. Despite the risks, Anthropic reports positive outcomes from the early phase of Project Glasswing. Since its inception, partners have collectively identified over 10,000 high or critical security flaws using the model. The company estimates that preventing a major cyber attack through these efforts could protect more than 100 million people. While major participants like Apple, Nvidia, Microsoft, CrowdStrike, and Palo Alto Networks have been publicly acknowledged, Anthropic has not disclosed the names of the 150 newly added companies joining the coalition.