Anthropic suspends Fable 5 worldwide after Trump export order

While the country celebrated a World Cup win and a Knicks championship, Anthropic spent the weekend in a fast-moving fight with the Trump administration over who gets to decide when a frontier AI model can stay in market. The company said a federal export control directive arrived at 5:21 p.m. ET on June 12 and forced it to cut off access to Fable 5 and Mythos 5 for all customers worldwide.

Anthropic said the order required it to suspend access for any foreign national, inside or outside the United States, including foreign national Anthropic employees. Because the company could not segment access that narrowly in practice, it disabled both models globally to comply. Access to all other Anthropic models was not affected. Anthropic said the government gave no specific details about its national-security concern, but officials appeared to believe they had found a way to bypass, or jailbreak, Fable 5’s safeguards.

That claim sits at the center of the dispute. Anthropic said its own review found only a small number of previously known, minor vulnerabilities, which it described as relatively simple and also present in other publicly available models. The company said it had spent thousands of hours red-teaming Fable 5 with the US government, the UK AISI, private third parties and internal teams, and that no tester had found a universal jailbreak. Anthropic also said perfect jailbreak resistance may not be possible.

The clash landed just days after Anthropic launched Fable 5 and Mythos 5 on June 9. The company said Fable 5 was its first broadly available Mythos-class model, state-of-the-art on nearly all tested benchmarks, with safeguard fallbacks triggered in less than 5% of sessions on average. It priced both models at $10 per million input tokens and $50 per million output tokens, and initially limited Mythos 5 to a small group of cyber defenders and infrastructure providers through Project Glasswing, a government collaboration and upgrade to Claude Mythos Preview.

The episode also sharpened a broader power struggle with Washington. On February 27, 2026, President Donald Trump directed federal agencies to stop using Anthropic technology, and on March 5 the Defense Department formally labeled Anthropic a supply-chain risk, a designation usually reserved for foreign firms tied to U.S. adversaries. Congressional Research Service said earlier negotiations with the Pentagon touched restrictions involving autonomous weapons and mass domestic surveillance.

That backdrop makes the latest confrontation more than a one-off product dispute. The White House reportedly ran a frantic 24-hour response involving Dario Amodei, Scott Bessent, Sean Cairncross, Susie Wiles and Andy Jassy. Amodei publicly backed mandatory testing and the option for government to block unsafe frontier models, but Anthropic said this directive did not follow the transparent, fair and technically grounded process it believes should govern such decisions. The precedent now being set is plain: future model launches may face not only safety reviews, but real-time federal limits on who can use them.