US Court Upholds Trump Administration Blacklist on Anthropic
A panel of judges ruled 2-1 in favour of government authority over AI military use.
The US Court of Appeals for the District of Columbia Circuit has upheld the Donald Trump administration's decision to blacklist technologies from artificial intelligence developer Anthropic. As reported by Ars Technica, a panel of judges ruled by a 2-1 majority that the government had full authority to restrict the use of the company's products because Anthropic refused to enable certain functions of Claude for military agencies.
Military risks and AI restrictions
The court's decision noted that the dispute touches on the issue of military deployment of new technologies with unprecedented capabilities. Two opposing positions emerged during the hearing.
On one side, the US government and Secretary of Defence Pete Hegseth argued that overly restricted AI models could suddenly shut down at a critical moment, potentially disrupting crucial military operations. On the other side, Anthropic pointed out that removing safety guardrails from an AI model could lead to it generating incorrect targets (hallucinating) and causing unintended consequences during combat.
Decision made within the law
The court explained that President Donald Trump and the Secretary of Defence were responsible for determining how to balance these competing risks. In the judges' assessment, the Secretary of Defence did not exceed the authority granted under the Supply Chain Security Act or the US Constitution.
Notably, the same court rejected an emergency motion by Anthropic to suspend the blacklisting back in April. This latest decision definitively dismisses the company's lawsuit, sending a clear signal across the tech industry as a whole.