Anthropic, the American AI company behind the Claude family of models, was forced to pull its newest and most powerful AI systems offline over the weekend after the White House demanded the company block access for all foreign nationals — including its own employees, according to The Verge.
The conflict centers on Fable 5, Anthropic's latest model, which government officials say was released without proper approval. "They screwed us," one administration official told The Decoder. The same outlet reports that Anthropic stands accused of disregarding a Trump administration cyber directive, and that talks are now underway involving the Department of Commerce, the CIA, and science advisor Michael Kratsios.
The Decoder also raises the question of whether Washington's demands are even technically achievable, suggesting the government may be pushing for "unhackable" large language models — a standard that experts consider impossible to meet in practice.
The international fallout has been swift. According to The Verge, the episode offered foreign observers a sober warning about the risks of relying on American AI infrastructure. Legal IT Insider notes the incident has exposed what it calls geopolitical risk in the AI supply chain, a concern now especially acute for law firms and enterprises that built workflows around Anthropic's tools.
The Baltimore Sun and WGXA both report the move has ignited a broader AI safety debate, with critics and supporters of government oversight clashing over how much control Washington should exercise over frontier AI systems.
For businesses and everyday users, the lesson is blunt: AI tools that seemed like permanent infrastructure can vanish overnight when governments intervene — and no one, apparently, is immune.