Anthropic has developed a new AI model called Claude Mythos, "Claude Mythos," which exceeds its predecessors' capabilities and is so powerful that its creator, Anthropic, is cautious about using it. The company has released ClaudeMythos Preview in a limited capacity due to concerns that hackers could exploit its advanced capabilities. This comes as the Department of Defense labels Anthropic as a national security supply chain risk, a designation Anthropic has been challenging in court. The meeting of the CEOs of America's largest banks was held at the Treasury Department to discuss the potential cyber risks posed by the Claude MythOS Preview model. JPMorgan, Apple, Google, Microsoft, and Nvidia have been enlisted in a cybersecurity initiative tied to the model, Project Glasswing. In November 2025, a Chinese state-sponsored group used Claude Code to execute the first known AI-orchestrated cyber espionage campaign, infiltrating 30 organizations, including tech companies, financial institutions, and government agencies.