Trump administration draws global backlash over Anthropic model order

The Trump admin went back to telling Anthropic about how its products can be used after the US government, on Friday, directed the AI firm to cut off foreign nationals from its newest Claude Fable 5 and Mythos 5 models.  

However, the “national security concerns” rationale did not go down well in many quarters, triggering immediate criticism from tech executives, policy analysts, and politicians outside the United States who warned that Washington is weaponizing AI access as a geopolitical tool.

Unlike earlier in the year when Anthropic resisted the government’s requests, this time around, the firm suspended access to both models for all customers while it works to comply with the directive. 

According to the company’s statement, the ban covers foreign nationals inside and outside the country. Even non-US citizens working for Anthropic did not get a pass.

Why is the US government trying to meddle with Anthropic again?

The June 12 export restriction is the latest installment in a months-long series of charged public interactions between the Trump administration and Anthropic. 

As Cryptopolitan reported earlier in the year, the Pentagon slapped Anthropic with a supply chain label over what the company described as resistance to greenlighting its AI tech to be used in domestic surveillance and fully autonomous weapons systems. 

That case dragged on until March, when a California federal judge ruled that the Trump admin violated free speech protections.

Now, barely three months later, the administration has moved on to a different front: export controls on Anthropic’s AI models themselves.

‘Cartoonish’ and ‘incoherent,’ critics say

Marc Andreessen, the venture capitalist, responded to the situation on X with a post criticizing regulation written by those who “never in his life built a single thing.”

Dean W. Ball, a senior fellow at the Foundation for American Innovation, questioned the “simply cartoonish” logic of an administration that has shown a willingness to hand advanced AI chips to China while throttling allied nations’ access to American AI models. 

Chris McGuire, a senior fellow at the Council on Foreign Relations, argued that the Commerce Department’s export control strategy “has been completely incoherent and sabotaging.” Washington is restricting US firms from releasing models even though it has no influence over how Chinese firms use the AI chips it approves for sale to China.

Peter Girnus, a senior threat researcher at the Zero Day Initiative, drew a comparison to the 1990s encryption wars, when the US government classified cryptographic software as a munition under arms export regulations. 

Girnus noted that those controls eventually collapsed. He also flagged a practical absurdity in the current order: the “deemed export” rule means Anthropic’s foreign-born employees who helped build the models are now locked out of them.

US allies and international observers wake up to AI sovereignty panic

Outside the US, the reaction was less philosophical. 

Muhammad Ziauddin “Zia” Yusuf, Reform UK’s spokesperson for home affairs, wrote on X that he had “warned for months that America would soon restrict access to state of the art frontier AI models for national security reasons,” adding that Britain has “virtually zero sovereign AI” capability as a result of domestic policy failures.

Indian entrepreneur Vasant Shetty also pointed out that India is the second-largest market globally for both ChatGPT and Anthropic’s products. “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,” Shetty wrote.

Advocates for local and open-source AI models also revived their arguments. Alex Finn, Founder and CEO of Henry Intelligent Machines PBC, wrote on X that “no company or government will ever be able to take away your local models.” 

How will Anthropic handle the ban on Fable 5? 

Anthropic launched Fable 5 and Mythos 5 on June 9. The restriction landed just days later. Investors must now factor in regulatory risks around the company’s core products ahead of its reported IPO, which could value the company near $1 trillion.

Dario Amodei, Anthropic’s CEO, foresaw possible regulatory headaches in his “Policy on the AI Exponential” June 10 essay, where he conceded that AI is advancing faster than legacy policy processes can handle.

The order represents a shift in how Washington approaches AI controls. Previous US export restrictions focused on chips and hardware. This is the first time the government has moved to restrict foreign access to the AI models themselves.

Dan Shipper, the CEO of Every, predicted the restriction would be lifted within days and that the net effect would be increased demand for Fable 5. But he acknowledged the disruption inside Anthropic, comparing it to the upheaval that followed Sam Altman’s brief firing from OpenAI in 2023.

If you're reading this, you’re already ahead. Stay there with our newsletter.