US court strikes down Anthropic supply chain risk designation, curbing executive sanctions against AI model makers

A US federal court in California has ruled unlawful the Trump administration's designation of Anthropic as a supply chain risk, marking a first legal victory for the Claude developer against Defense Department efforts to restrict its business across government agencies.

In her ruling, US District Judge Rita Lin stated that the decision led by Defense Secretary Pete Hegseth to designate the company a national security threat constituted unlawful retaliation in violation of the First Amendment, calling the move arbitrary and capricious and noting that it deprived the company of Fifth Amendment due process after the administration directed all federal agencies earlier this year, including non-military bodies, to cease dealing with the Claude maker.

The dispute stemmed from Anthropic's insistence on strict safety controls barring the deployment of its models in fully autonomous weapons or mass surveillance, which the Pentagon viewed as an attempt to impose guardianship over paid model use.However, the court found that the administration's conduct and public remarks demonstrated an intent to make a public example of the company for what officials described as its audacity in criticizing government policy, pointing to core contradictions in the Pentagon's stance, which simultaneously sought to invoke the Defense Production Act against Anthropic as a critical national security asset while continuing procurement efforts and collaboration with its new Mythos model on cybersecurity missions.

The ruling dismissed the Pentagon's technical concerns, confirming that Anthropic retains no backdoor access to its systems once delivered to government entities, with the judge emphasizing that defense bodies' discretion in selecting AI vendors does not grant a blank check to exploit national security pretexts to punish vendors over policy disagreements, as a second lawsuit filed by the company in Washington DC since last March proceeds.

The decision carries direct practical implications for technology and government sectors in the Gulf, Egypt, and the Levant that rely on Claude models or are evaluating their integration into enterprise cloud architectures, easing compliance concerns and sudden geopolitical risks that threatened supply continuity and API access under unilateral executive actions. The dispute also provides technology leaders across the region with a clear benchmark on licensing terms, confirming that a model provider's insistence on ethical use policies does not imply backdoor control mechanisms that undermine data sovereignty or system autonomy in local deployments.