Anthropic's two most powerful AI models are coming back online after a standoff with the U.S. government.
The company said the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5, and that it would begin restoring access on Wednesday, July 1. Anthropic announced the news in a post on X.
According to CNBC, the U.S. Department of Commerce lifted the controls on both models. The Verge reports the reversal came after weeks of negotiating with the Trump administration, a process Business Insider and others described as talks with the White House.
The restrictions were not longstanding. According to CNBC-TV18, the government halted access to the models on June 12 over national security concerns, prompting Anthropic to suspend them. Fox Business and the New York Post framed the reversal as the end of a bitter standoff, while FOX 8 tied the original restrictions to a cybersecurity alarm.
The restored access arrives alongside other product news. Cybernews reports Anthropic is streamlining research with a new tool called Claude Science and rolling out a cheaper Sonnet 5 model. Forbes described the new workbench as mapping a writer's field for $26.
Not all the coverage is celebratory. Fortune argues that even with Fable back, U.S. AI policy remains a mess. And some outlets flagged early performance questions: one report cited an AI coding group warning of a Fable 5 performance collapse after the relaunch, while another pointed to a leak suggesting a rival model outperformed it.
Why it matters: government export controls can switch a leading AI product on or off almost overnight, showing how much the tools businesses and researchers rely on now depend on shifting national-security policy.