The United States government has moved on two fronts against Anthropic, the AI company behind the Claude family of models, in actions that mark one of the most sweeping government interventions in the commercial AI sector to date.
According to reporting by DQ India, President Donald Trump directed every federal agency to immediately stop using Anthropic's AI models. Agencies including the Department of Defense are being given a six-month window to complete the phase-out, following what the outlet described as a "Pentagon AI showdown."
Separately, a US export control directive has barred foreign nationals from accessing two of Anthropic's newest models, Fable 5 and Mythos 5, citing AI safety concerns. According to Business Standard, Anthropic responded by disabling both models globally — meaning users outside the United States lost access as a direct consequence of a domestic regulatory order.
The dual actions illustrate a dynamic that analysts and commentators have warned about for some time. An opinion piece published by NDTV argues that the episode validates the case for "sovereign AI" — the idea that nations and businesses relying on externally controlled AI systems are exposed to sudden, unilateral cutoffs. As the piece puts it: "if you don't own the model, you don't own your access to it."
For governments, companies, and developers worldwide that have built workflows or products on top of Anthropic's services, the lesson is stark: a single policy decision in Washington can instantly reshape what tools are available to them, with little warning and no appeal.