Anthropic got hit by export rules nobody understands

The Trump administration ordered Anthropic to cut access for all foreign nationals, including users inside the US and its own employees, to its AI models Fable 5 and Mythos 5, citing national security authorities under export control rules [1]. Anthropic stated on its website that the government cited "national security authorities" to justify "an export control directive" on the models [1]. The company also claimed that government concerns about a "jailbreak" potentially used by groups linked to China did not allow users to circumvent all of its safeguards [1]. The models remain hosted on Anthropic's servers, and users do not receive source code or model weights [1]. Hanna Dohmen, a senior research analyst at Georgetown University's Center for Security and Emerging Technology, said, "To my knowledge, this is the first time US export controls have been used to control access to an AI model in this way" [1]. The legal basis for the order has not been publicly explained by the administration [1]. Export controls have traditionally applied to tangible goods that can be shipped across borders, later expanding to cover software, source code, and technical data [1]. Remote access to cloud services remains a known gap in current export control regimes, one that Congress is already trying to close through legislation [1]. Andrew Reddie, a professor at UC Berkeley's Goldman School of Public Policy, said, "To say that this is an unsettled area of export control rule-making would be an understatement" [1]. He noted that export control rules and arms regulations give the government "wide latitude" to restrict access to certain goods, but "the equivocation by successive administrations regarding what the responsibilities of model developers are" has made compliance difficult for firms [1]. Anthropic, founded in 2021 by former OpenAI members, had an estimated valuation of $965 billion in May 2026, making it the most valuable pure-play AI company in the world [8]. The company has been in conflict with the Department of Defense since January 2026 over the use of its products for military purposes and mass domestic surveillance [9]. US federal agencies began phasing out use of Claude after Anthropic refused to remove contractual prohibitions on those uses, and the DoD designated the company a "supply chain risk" [10]. A federal judge issued a temporary injunction against that designation on March 26, 2026 [10]. Reddie said the episode "makes clear the unsustainability of the existing governance regime" [1]. He added that if creating models impossible to jailbreak becomes the de facto standard, "then it will have no AI models" [1]. The incident has raised questions about whether governments and companies outside the US should rely on American firms for access to strategically important systems [1].