Anthropic loses bid to pause Pentagon blacklisting as AI legal battle escalates

Serving tech enthusiasts for over 25 years.

TechSpot means tech analysis and advice you can trust.

What just happened? Anthropic has suffered another setback in its escalating fight with the Pentagon. A federal appeals court in Washington, D.C., has refused to temporarily block the Defense Department from treating the AI firm as a supply chain risk while the case continues.

The ruling keeps Anthropic locked out of DoD contracts for now, even though a separate federal court in California recently barred the Trump administration from enforcing a broader ban on the use of Claude.

The split decisions leave Anthropic in an awkward position. It's still able to work with other government agencies, but effectively frozen out of Pentagon-related business.

In its decision, the appeals court said the balance of harms favored the government. Judges described Anthropic's risk as primarily financial, while casting the Pentagon's position as part of wartime procurement during an active military conflict.

Also read: Anthropic unveils new powerful AI that finds software flaws, but says it's too dangerous to release

The panel acknowledged the company would likely suffer some irreparable harm without relief, but said Anthropic had not shown that its free speech rights were being meaningfully "chilled" while the case plays out.

The court did, however, say the matter should be expedited, suggesting the legal fight will move faster because of the damage Anthropic is likely to face in the meantime.

Anthropic CEO Dario Amodei

The dispute stems from a breakdown in negotiations between Anthropic and the Pentagon over how Claude could be used in defense settings.

Anthropic had pushed for assurances that its models would not be used for fully autonomous weapons or domestic mass surveillance, while the DoD wanted broader access across all lawful uses.

When the two sides failed to reach an agreement, Defense Secretary Pete Hegseth publicly branded the company a supply chain risk before the department formalized the designation in early March.

Defense contractors are now required to certify they are not using Anthropic's models in work tied to the military, though Claude remains available for non-DoD projects and other federal agencies. Anthropic has argued the move is retaliatory, unconstitutional, arbitrary, and procedurally unsound.

Part of the confusion comes from the government relying on two separate legal authorities to justify its actions, forcing Anthropic to challenge them in different courts. That explains how the California court could block one aspect of the administration's restrictions while the Washington appeals court allowed the Pentagon blacklist to remain in place.

Anthropic is reportedly the first American company to receive this kind of supply chain risk designation, a label more commonly associated with foreign adversaries. It's a dramatic turn of events for a company that signed a $200 million Pentagon contract last July and was once being positioned for deeper integration into the DoD's AI efforts.