The first forced removal of a widely-deployed frontier model by government export controls signals that high-capability AI availability is now a managed policy variable, not a given. On June 12, 2026, the U.S. government ordered Anthropic to suspend access to Fable 5 and Mythos 5 for all foreign nationals, citing a reported jailbreak. Anthropic disabled both models for all customers to comply, calling the action a misunderstanding - the flagged vulnerability was narrow, non-universal, and Anthropic said capabilities were already available via other deployed models including GPT-5.5. Axios reported June 27 that Fable 5, offline for 15 days, is on track to return within days after negotiations progressed. Commerce Secretary Howard Lutnick's June 26 letter, first reported by Semafor, cleared Mythos 5 for roughly 100 U.S. institutions. Anthropic International Managing Director Chris Ciauri stated at a Seoul press conference: "We are very confident that in the coming days, the models will become available again." Pentagon and NSA sign-off on Fable 5 remains outstanding as of June 28.
Policy Pulse – Issue #21 – Week of June 27, 2026