The Trump administration ordered Anthropic to restrict its most advanced cybersecurity artificial intelligence (AI) model on June 13, 2026, prompting the company to suspend access worldwide. Anthropic shut down global access to two AI models, Mythos and Fable 5, following a Commerce Department directive citing national security concerns. Semafor reports that a China-linked group was suspected of gaining unauthorized entry to the system, partly motivating the export control decision. U.S. officials feared Beijing could reverse-engineer Mythos through a process known as “distillation,” effectively replicating its capabilities. Anthropic launched Claude Mythos Preview in April 2026 as a frontier AI model built around autonomous cybersecurity offense and defense, with capabilities including autonomous identification of thousands of high-severity zero-day vulnerabilities and 83.1% accuracy on the CyberGym vulnerability reproduction benchmark. Access was restricted to roughly 12 named corporate partners and 40-plus additional organizations through Project Glasswing. Fable 5, a consumer-facing version of Mythos released in early June, was reportedly jailbroken; White House AI adviser David Sacks posted that Anthropic CEO Dario Amodei was notified but reportedly downplayed the risk and declined to address it.