Anthropic has expanded Project Glasswing, its collaborative initiative to identify and remediate critical software vulnerabilities using artificial intelligence, to approximately 150 additional organizations across more than 15 countries. The announcement follows the company’s recent confidential filing for an initial public offering, which comes on the heels of a 65 billion dollar funding round that valued the firm at nearly one trillion dollars. Central to this expansion is Claude Mythos, Anthropic’s latest language model, which the company describes as its most capable version to date. The model demonstrated the ability to detect thousands of zero-day vulnerabilities across multiple weeks of testing. Anthropic initially granted early access to Claude Mythos Preview in early April, providing 50 partner organizations, including U.S. government entities, with scanning capabilities. The newly expanded cohort now includes entities operating within the power, water, healthcare, telecommunications, and hardware sectors, filling a gap in the initial program’s sector representation. According to Anthropic, the participating organizations manage critical codebases that serve as foundational infrastructure for other industries and governments. The company estimates that a successful compromise of any single partner’s systems could disrupt services for over 100 million individuals, posing significant threats to global and national security. The international roster includes organizations from the United States, Australia, Canada, France, Germany, Italy, Switzerland, the Netherlands, Spain, Belgium, Sweden, India, Japan, New Zealand, and South Korea. Notable participants identified in reporting include cybersecurity platform Okta, South Korean technology firms Samsung, SK Hynix, and SK Telecom, NATO, and the European Union’s Agency for Cybersecurity, ENISA. Anthropic emphasized the urgency of scaling Project Glasswing, noting that the broader artificial intelligence industry is rapidly closing the capability gap. Competitors are expected to launch models with comparable cybersecurity detection capabilities in the near term, prompting Anthropic to establish defensive safeguards and industry standards before rival systems achieve similar maturity. The initiative reflects a growing trend of integrating large language models directly into critical infrastructure defense, prioritizing proactive vulnerability management over reactive patching. The expanded deployment marks a strategic pivot toward securing the global technology supply chain. By distributing advanced AI scanning tools to high-risk sectors and allied nations, Anthropic aims to standardize automated threat detection while mitigating cascading failures across interconnected digital ecosystems. As the artificial intelligence race intensifies, Project Glasswing positions Anthropic at the forefront of applying generative models to enterprise-grade cybersecurity operations.