Anthropic disabled access to its two most capable models, Claude Fable 5 and Mythos 5, late on Friday after the US Commerce Department issued a national-security export-control directive barring any foreign national from using them — including Anthropic's own non-citizen employees inside the United States.
What the directive does
Commerce Secretary Howard Lutnick sent the letter to CEO Dario Amodei at 5:21 pm Eastern Time on June 12, 2026, three days after Fable 5 and Mythos 5 launched. Because Anthropic cannot filter foreign nationals from US users in real time, the company shut the two Mythos-class models down for every customer. Access to Claude Opus 4.8 and the rest of the Claude family is unaffected.
The jailbreak Anthropic disputes
Anthropic said officials cited a technique to bypass Fable 5's safeguards — three layers of classifiers that route cybersecurity, biology and chemistry, and distillation queries to Opus 4.8 in under 5% of sessions. The company called the demonstration narrow rather than universal, noted that other public frontier models including OpenAI's GPT-5.5 would produce the same outputs without any bypass, and warned the standard could halt frontier deployments industry-wide.
The political backdrop
The Pentagon labeled Anthropic a "supply chain risk" in March after the company refused contract terms permitting use "for any lawful purpose," seeking carve-outs for autonomous weapons and mass domestic surveillance. Anthropic is challenging that designation in federal court. Recent advisors including David Sacks and Emil Michael have publicly attacked the company. Anthropic launched Fable 5 and Mythos 5 just days earlier as its first publicly available Mythos-class model.
Customer fallout
Stripe had used Fable 5 to compress a 50-million-line Ruby migration into days; Hebbia ranked it top on its finance benchmark; Mozilla used Mythos-class capabilities to resolve hundreds of software vulnerabilities through Project Glasswing. Those workflows are paused while Anthropic disputes the order, which the company called "a misunderstanding." The episode arrives weeks after the company confidentially filed for an IPO at a $965 billion valuation, and against the backdrop of Anthropic's SpaceX compute partnership bringing online 300 megawatts of Colossus 1 capacity.
Reporting based on coverage from Fortune, MarkTechPost, Bloomberg, Axios, CNBC and Anthropic's official statement.