Anthropic PBC filed two lawsuits on Monday against the Trump administration, seeking to block a Pentagon designation that labels the AI company a supply-chain risk to national security. The San Francisco-based firm, maker of the Claude AI model, accused federal agencies of unlawful retaliation for maintaining safeguards against using its technology in lethal autonomous weapons and mass surveillance of Americans.

One lawsuit landed in U.S. District Court for the Northern District of California (Case No. 3:26-cv-01996), naming the Department of Defense—referred to as the Department of War in the filing—along with more than a dozen other agencies and officials including Defense Secretary Pete Hegseth, Treasury Secretary Scott Bessent and Commerce Secretary Howard Lutnick. A second suit went to the U.S. Court of Appeals for the D.C. Circuit. Anthropic seeks declaratory and injunctive relief to halt enforcement of the designation, a presidential directive and related agency actions.

The dispute arose from months of negotiations between Anthropic and the Pentagon over usage policies for Claude, which the Defense Department had used extensively in classified systems for intelligence and operations, including in Iran. Anthropic developed specialized "Claude Gov" models and secured partnerships worth up to $200 million, but refused Pentagon demands for "all lawful use," citing untested risks in autonomous warfare without human oversight and surveillance of U.S. citizens.

Tensions peaked in late February. On February 24, CEO Dario Amodei met with Hegseth, who issued an ultimatum for compliance by February 27. Amodei publicly stated the next day that Anthropic could not drop the safeguards but offered a smooth transition to another provider. On February 27, President Trump issued a directive ordering all federal agencies to cease using Anthropic products, and Hegseth designated the company a supply-chain risk—the first time for a U.S. firm, typically reserved for foreign adversaries.

Agencies quickly followed. The General Services Administration terminated contracts, Treasury and State departments switched providers, and others like Health and Human Services disabled Claude. Hegseth's March 4 letter formalized the exclusion from defense contracts and partners. Trump gave six months for transitioning from classified systems.

Anthropic alleges First Amendment violations for punishing protected speech on AI safety, Administrative Procedure Act breaches for arbitrary actions exceeding authority under 10 U.S.C. § 3252, ultra vires presidential overreach and Fifth Amendment due process denials. The company claims irreparable harm to its reputation, contracts worth hundreds of millions and partnerships like Palantir.

"Anthropic turns to the judiciary as a last resort to vindicate its rights and halt the Executive’s unlawful campaign of retaliation," the company stated. White House spokesperson Liz Huston responded that Trump "will never allow a radical left, woke company to jeopardize our national security by dictating how the greatest and most powerful military in the world operates." The Pentagon declined comment on litigation.

The case could reshape government-AI company relations, testing executive power over supply chains and ethical restrictions amid rising military AI demands. Anthropic projects $14 billion in 2026 revenue, much from enterprise and government-linked deals.