Court Upholds Pentagon Ban on Anthropic AI Models

Overview

A federal appeals court in Washington, D.C., has upheld the Pentagon’s decision to blacklist Anthropic, dealing a significant legal blow to the artificial intelligence company. The 2-1 ruling rejected Anthropic’s argument that the Department of Defense’s ban on its Claude models was arbitrary, unauthorized, and unconstitutional.

The designation, originally issued in March, classifies Anthropic as a supply chain risk, effectively preventing the U.S. military from using its AI models and blocking defense contractors from integrating them into work for the agency. This decision marks a pivotal moment in the escalating legal and political conflict between the tech firm and the Trump administration.

"The Department had ample support for its conclusion that the continued integration of Claude into the Department's information systems... presented a statutorily covered national-security risk." — Judge Gregory Katsas

Legal Ruling and Judicial Breakdown

The U.S. Court of Appeals for the District of Columbia issued its decision on Friday, with the majority opinion written by Judge Gregory Katsas. Judge Neomi Rao, who was appointed by President Donald Trump, joined the majority. Both judges affirmed that the government’s assessment of national security threats posed by Anthropic’s technology was well-supported.

The court found that the Pentagon acted within its statutory authority to protect sensitive information systems from potential vulnerabilities introduced by foreign or untrusted AI vendors. The majority opinion emphasized that the Defense Department’s concerns were not speculative but rooted in concrete risks regarding the integration of proprietary models into military infrastructure.

However, the panel was not unanimous. Judge Karen LeCraft Henderson, appointed by former President George H.W. Bush, filed a dissenting opinion. While the specifics of her dissent were not detailed in the immediate release, her disagreement highlights the complex legal and ethical questions surrounding government restrictions on private AI development.

Previous Legal Challenges

This ruling follows a parallel legal battle in San Francisco. In March, Anthropic sued the Trump administration in both U.S. District Court in San Francisco and the D.C. Circuit Appeals Court to reverse its blacklisting. The Department of Defense relied on two distinct designations to justify its supply chain risk action, requiring litigation in two separate courts.

Just last month, a federal judge in San Francisco ruled that one of those designations was illegal. The Friday decision by the D.C. appeals court upheld the second designation, leaving Anthropic still barred from government contracts despite the partial victory in California.

An Anthropic spokesperson responded to the ruling with stated confidence in their legal position. "We respectfully disagree with the court's decision," the spokesperson said. "Another federal court has already held the government's parallel designation unlawful. We remain confident in our position and are considering all options, including further review."

Origins of the Conflict

The rift between Anthropic and the Trump administration began earlier this year after negotiations regarding the deployment of Claude on the Pentagon’s GenAI.mil platform collapsed. Anthropic had signed a $200 million contract with the Pentagon in July 2025, but talks broke down in September when the two sides could not agree on usage terms.

The core dispute centered on control and ethical boundaries. The Defense Department sought to grant the military unfettered access to Anthropic’s models for all lawful purposes. In contrast, Anthropic demanded assurances that its technology would not be used for fully autonomous weapons or domestic mass surveillance.

Defense Secretary Pete Hegseth accused Anthropic of attempting to seize veto power over the operational decisions of the United States military. He argued that the company was trying to dictate how the U.S. military could use tools it had paid to develop.

Judge Katsas echoed these concerns in his opinion, writing that the Department reasonably feared Anthropic might manipulate Claude’s design to prevent it from performing national-security functions deemed necessary by the government.

Political Fallout and Future Options

The legal battle has intensified a broader political feud between Anthropic CEO Dario Amodei and President Trump. Amodei recently drew the President’s ire by calling for an industry-wide slowdown in AI development. In response, Trump has repeatedly criticized Amodei on social media, accusing him of pretending to be a "perfect little angel" after previously raising concerns about AI risks.

The personal nature of the dispute was further highlighted when Amodei was not invited to the recent state dinner hosted by the White House for Chinese President Xi Jinping. Trump took to Truth Social to vent his frustration, writing that the administration would continue to stop "people" like Amodei from hindering progress.

Despite the setback, Anthropic has several remaining avenues to challenge the ruling. The appellate panel indicated it would delay the decision from taking immediate effect, giving the company time to petition the same panel for a rehearing. Additionally, Anthropic could seek an en banc rehearing, which would involve all judges on the D.C. Circuit Court of Appeals.

If those efforts fail, the company may appeal directly to the Supreme Court. A hearing at the highest level could take months or even years, during which the status quo of the blacklisting will remain in place.

Meanwhile, the implications for the defense AI sector are significant. The ruling reinforces the government’s stance that national security considerations can override commercial partnerships and intellectual property concerns. It also serves as a stark warning to other tech companies navigating relationships with the military: alignment with federal security priorities may be a prerequisite for accessing defense contracts.