US blocks Anthropic's $900B AI models in sweeping export order

The US government ordered Anthropic to block all foreign nationals from accessing its Fable 5 and Mythos 5 AI models, citing national security concerns over a potential jailbreak vulnerability, the company said Friday.

"We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people," Anthropic said in a blog post. "If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers."

The directive, issued by the Commerce Department, bars any foreign national from using the models "whether inside or outside the United States, including foreign national Anthropic employees," the company said. Anthropic shut off access to both systems for all customers to comply. The order came after researchers at Amazon.com Inc. identified a method to bypass Fable 5's safety guardrails, according to a person familiar with the matter. Amazon Chief Executive Andy Jassy was involved in exchanges with US officials before the controls were imposed, one of the people said.

The move threatens to set a precedent for the entire AI industry as startups including Anthropic — valued at more than $900 billion — and OpenAI pursue initial public offerings. It also escalates tensions between Anthropic and the Trump administration, with the company already suing the government over a supply chain blacklist imposed after it refused to let the US military use its models for domestic surveillance and fully autonomous weapons systems.

A Precedent for Silicon Valley

The export control directive marks a sharp reversal from the Trump administration's earlier stance. A June executive order declared the US would not create a mandatory licensing regime for AI models, yet Friday's action effectively does exactly that. "For anyone who was naive and perhaps hoping that this leverage wouldn't be exerted, it's a massive wake-up call," Aidan Gomez, co-founder of Nvidia Corp.-backed AI startup Cohere Inc., said in an interview.

The order has drawn criticism from across the tech sector. Dean Ball, a former White House AI adviser who helped author the administration's AI Action Plan, called it "baffling" that the US would restrict foreign access to Anthropic models while simultaneously allowing exports of advanced AI chips to China. "I can't tell if this is lawfare against Anthropic in particular or extreme national-security hawkery," he wrote on X.

Gary Marcus, a scientist known for his work on AI, called the directive "shocking" and said it would prove "counterproductive for the US AI industry." He added that Chinese nationals working at US AI companies "will consider returning to the competition in China ASAP."

Global Fallout and G-7 Diplomacy

The ban has alarmed foreign governments and industry executives who fear unilateral US actions could dictate who can use leading AI tools. Ciaran Martin, former chief of the UK's National Cyber Security Centre, said the order "came completely out of the blue" and is "probably going to prove unsustainable" because the only practical way to comply is to withdraw the model entirely. "The US is obsessed with winning the AI race with China," he added. "This slows it down."

Vassant Shetty, an Indian entrepreneur, said the move is a "huge wake-up call" for countries like India, the second-largest market for both ChatGPT and Anthropic. "If they can turn off the access at the press of a button like this, we are absolutely at the mercy of a foreign govt," he posted on X.

At the Group of Seven summit in Evian-les-Bains, France, President Trump said negotiations with Anthropic were "going fine" after a working lunch that included Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman and Google DeepMind CEO Demis Hassabis. Treasury Secretary Scott Bessent, one of the officials who targeted Anthropic's models, also attended. Anthropic said it is working to restore access and believes the directive stems from a "misunderstanding."

The European Union's executive arm said it is assessing the situation and continuing to talk to allies about potential risks, adding that the developments underline Europe's need for technological sovereignty.

This article is for informational purposes only and does not constitute investment advice.