Anthropic has disabled access to its latest AI models Fable 5 and Mythos 5 following a U.S. government export control directive citing national security. The company expressed regret, calling for a transparent and fair process.
Anthropic disabled access to its latest AI models, Fable 5 and Mythos 5, on April 18, 2026, following a U.S. government export control directive citing national security. The order prohibits access by any foreign national, including Anthropic's own foreign employees.
Fable 5 and Mythos 5 are cutting-edge models recently unveiled by Anthropic, with Fable 5 being the first advanced model released to the public with new safeguards blocking responses in high-risk areas. The move comes shortly after Anthropic demonstrated its cybersecurity capabilities with Claude Mythos Preview. The company is already in litigation with the Department of Defense over a supply chain risk designation.
This incident highlights tensions between AI model export controls and national security. Anthropic supports the government's ability to block unsafe deployments but criticizes the lack of transparency and fairness in this action. It sets a precedent for AI regulation and corporate response.
Some users criticized Anthropic for blocking model access at the government's request, comparing it to Google's quick abandonment of its 'Don't be evil' slogan. This reflects skepticism about how easily AI companies' ethical commitments can waver under government pressure.