A federal court has struck down the Trump administration’s attempt to blacklist Anthropic from government contracting, ruling that designating the AI company as a “supply chain risk” to national security was both unlawful and retaliatory.
Judge Rita F. Lin of the US District Court in Northern California granted partial summary judgment in Anthropic’s favor on August 27, calling the designation “arbitrary and capricious” and a violation of the company’s First Amendment rights. The 59-page opinion effectively restores Anthropic’s ability to compete for federal contracts worth potentially billions of dollars.
How an AI ethics dispute became a national security fight
The conflict traces back to negotiations over a roughly $200M contract for Anthropic’s AI models. During those talks, Anthropic insisted on including ethical usage guidelines that would restrict certain military applications of its technology. President Trump and Defense Secretary Pete Hegseth publicly accused the company of endangering national security by refusing to let the military use its AI without guardrails. On February 27, 2026, the administration issued a formal supply chain risk designation against Anthropic.
That designation is a serious weapon in the government’s regulatory arsenal. It’s the kind of label typically reserved for foreign adversaries like Huawei or Kaspersky, companies suspected of building backdoors for hostile governments. The practical effect was immediate and severe. Federal agencies across the board, including the Department of Defense, Treasury, and Homeland Security, were barred from doing business with Anthropic. The company claimed the blacklisting threatened billions in potential revenue and inflicted serious reputational damage.
Anthropic filed suit on March 9, 2026, and secured a preliminary injunction just 17 days later on March 26, suggesting the court saw problems with the government’s case from the start.
The court’s reasoning
Judge Lin found that the designation lacked any legitimate factual foundation. The administration couldn’t point to evidence that Anthropic’s technology posed a sabotage risk or that its supply chain was compromised. Instead, the record suggested the designation was punishment for Anthropic’s refusal to drop its ethical guidelines during contract negotiations.
The court concluded that the government had retaliated against Anthropic for exercising its right to set terms for how its products are used, and that punishing that speech with a national security blacklist crosses a constitutional line.
While the decision blocks multiple federal agencies from enforcing the blacklist, it does not compel the Pentagon to actually buy Anthropic’s products. The military retains full discretion over its procurement decisions.
What this means for AI and government contracting
This is the first major court test of whether the government can weaponize supply chain security designations against domestic companies for policy disagreements rather than genuine security concerns. The answer, at least from this court, is a clear no.
AI firms that want to impose usage restrictions on their models now have a legal backstop: the government can choose not to buy your product, but it can’t blacklist you for having terms it doesn’t like.