Anthropic pulls plug on new AI models after Trump admin directive

Anthropic announced Friday it will temporarily remove access to its latest AI models, Fable 5 and Mythos 5, in response to a directive from the Trump administration. The order, which cites security concerns, restricts foreign nationals from using the models, though the government has not disclosed specifics such as the duration of the restrictions. Anthropic stated the directive appears to target a method to bypass its Fable 5 model, identifying minor vulnerabilities that other publicly available models can also detect. The company had previously released Mythos in April to about 40 organizations managing critical infrastructure, later expanding access to 150 groups. Fable 5, introduced Tuesday, is a more restricted version of Mythos with additional safeguards against cybersecurity threats. The White House has previously raised concerns about powerful AI models, with President Trump signing an executive order in June requiring companies to allow government oversight before public deployment. Anthropic CEO Dario Amodei supported the president’s directive, arguing governments should block unsafe AI deployments. However, the company criticized the current order, stating it lacks transparency, fairness, and technical grounding. This is not the first conflict between Anthropic and the Trump administration; earlier this year, Defense Secretary Pete Hegseth designated Anthropic a supply chain risk, prohibiting use of its Claude AI model. Anthropic sued the administration, warning the move could set a harmful precedent for U.S. companies. In its Friday statement, Anthropic expressed regret for the disruption, calling the restrictions a misunderstanding. The company assured it is working to restore access as soon as possible while emphasizing its belief in fair, transparent oversight for AI safety standards.