US Commerce Department orders Anthropic to suspend AI model exports

The US Commerce Department just did something it has never done before: it ordered a specific AI company to pull the plug on its most advanced models worldwide.

Anthropic received the directive on June 12 at 5:21 p.m. ET, commanding the company to suspend access to its Fable 5 and Mythos 5 models for all foreign nationals. The kicker: the order extends even to Anthropic’s own US-based employees who hold foreign citizenship, thanks to something called “deemed export” rules. In English: if you’re a non-US citizen working at Anthropic’s San Francisco office, you can’t touch the models either.

Anthropic complied immediately, disabling access to both models globally. The models had been live for exactly three days, having launched on June 9.

Why the government stepped in

The stated justification is national security. The directive reportedly traces back to a jailbreak demonstration that exposed vulnerabilities in Anthropic’s models, with the exploited capabilities said to be comparable to functions found in other frontier models like OpenAI’s GPT-5.5.

The scope of the order is remarkably broad. Export controls on technology typically target specific countries or entities. This directive applies to all foreign nationals, everywhere.

The “deemed export” wrinkle

The deemed export rule deserves its own explanation because it’s the detail that makes this order particularly disruptive. Under US export control law, sharing controlled technology with a foreign national inside the United States counts as an “export” to that person’s home country. It’s a rule that has existed for decades in contexts like defense technology and nuclear research.

Applying it to an AI model is different. AI companies employ thousands of researchers from dozens of countries. If Fable 5 and Mythos 5 are now classified under export controls strict enough to trigger deemed export restrictions, Anthropic potentially has to restrict internal access for a significant portion of its own workforce.

Industry observers have warned that regulatory actions of this nature could significantly hinder the deployment of frontier AI models across the sector.

Crypto markets see an opening

While traditional AI companies grapple with the fallout, a corner of the crypto market is having a very different kind of day. Tokens associated with decentralized AI projects saw immediate price increases following the announcement.

Venice (VVV) and Morpheus (MOR) both rallied as traders latched onto a straightforward narrative: if governments can flip a switch and disable centralized AI models, then decentralized alternatives that operate outside direct government control become more attractive.

Decentralized AI platforms distribute model inference and sometimes training across networks of independent nodes. There’s no single company to serve with a compliance order. Whether these platforms can actually match the capability of frontier models from Anthropic or OpenAI is a separate question.

What investors should watch closely is whether other AI labs receive similar orders in the coming weeks. If Anthropic’s competitors face the same restrictions, the entire frontier AI sector gets repriced, and the decentralized alternative narrative strengthens considerably. If Anthropic stands alone, this looks more like a targeted enforcement action than a paradigm shift.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.