Federal appeals court agrees with Pentagon: Anthropic is a 'national security risk' for America

A federal appeals court in Washington, D.C has sided with the Pentagon, affirming its designation of Anthropic as a ‘national security risk’. The 2-1 ruling blocks the AI company’s models from being used by the US military and its contractors, marking a significant setback for Anthropic in its ongoing battle against the Trump administration’s blacklisting. According to a report by CNBC, the majority opinion, delivered by Judges Gregory Katsas and Neomi Rao, concluded that the Department of Defense had “ample support” for its determination that Anthropic’s Claude models posed operational risks. Judge Karen LeCraft Henderson dissented, arguing the designation was overly broad.Anthropic's AI products pose a supply chain riskThe US Court of Appeals for the District of Columbia Circuit ruled 2-1 that the Defense Department had good reason to conclude that Anthropic's AI products pose a supply chain risk to the department and its contractors, because the company builds code into its models restricting certain tasks. Writing for the majority, Judge Gregory Katsas said the Department of Defense had adequately demonstrated that keeping Claude embedded in its information systems created a risk to national security covered by statute. The court also rejected Anthropic's argument that the Pentagon's action was retaliatory, tied to the company's stance on AI safety and ethics rather than legitimate security concerns.The ruling expands the power of the executive branch to blacklist U.S. businesses, and stems directly from Anthropic's refusal to loosen restrictions it maintains on how its Claude models can be used — specifically its stated policy against allowing its AI to be used for autonomous weapons or mass domestic surveillanceHow the standoff between Anthropic and Trump administration beganThe conflict between Anthropic and the Trump administration grew out of a $200 million contract dispute over how the Pentagon could deploy Claude on classified systems. The Defense Department demanded Anthropic make its models available for any lawful military use without restriction, while Anthropic sought written guarantees that Claude would not be deployed in fully autonomous weapons systems.On February 27, the Pentagon gave Anthropic a 5 p.m. deadline to comply with its demands that the government be able to use the service as it saw fit. Before that deadline passed, President Trump announced that no government workers would be allowed to use Anthropic, and Defense Secretary Pete Hegseth subsequently labeled the company a supply-chain risk, a designation that effectively blocks Anthropic from any government contracts and is more typically reserved for companies tied to foreign adversaries.A split legal battlefieldAnthropic sued the government over the designation on March 9, arguing the Pentagon had applied it inappropriately to punish the company on ideological grounds. That case unfolded in two separate courts: the DOD relied on two distinct designations to justify its action, which meant they had to be litigated separately. A San Francisco federal judge ruled last month that one of those designations was illegal — U.S. District Judge Rita Lin officially ruled on August 28 that the designation was illegal and that Hegseth had exceeded his authority. But Friday's ruling from the D.C. appeals court upheld the second, separate designation, leaving Anthropic's Pentagon blacklisting effectively intact despite its earlier win in San Francisco.Anthropic's response"We respectfully disagree with the court's decision," an Anthropic spokesperson said in a statement, adding that "another federal court has already held the government's parallel designation unlawful." The company said it remains confident in its position and is considering all options, including further review. The appellate panel said it would delay the decision from taking immediate effect, giving Anthropic time to petition for a rehearing before the same panel, seek an en banc rehearing before the full D.C. Circuit, or potentially ask the Supreme Court to take up the case.You use AI every day. Now get your AI Quotient. Take the AIQ test.