Anthropic, an AI company backed by Amazon, has sued the U.S. government, alleging retaliation for its refusal to remove safety limits on its Claude AI model. The dispute centers on the government's demand for "all lawful uses" of Claude, which Anthropic opposed for lethal autonomous warfare and mass surveillance of Americans.
Anthropic claims it developed a specialized "Claude Gov" version for national security work, but negotiations faltered when the Department of Defense insisted on an unrestricted usage policy. The company stated it was willing to work with the military, but not under terms compromising its ethical guidelines.
According to Anthropic's lawsuit, Secretary of Defense Pete Hegseth presented an ultimatum: comply within four days or face being designated a national security risk and expelled from the defense supply chain. Following Anthropic CEO Dario Amodei's public rejection, President Donald Trump ordered federal agencies to cease all use of Anthropic's technology, labeling the company "RADICAL LEFT, WOKE COMPANY."
The government subsequently terminated Anthropic's government-wide contract, citing supply-chain risk. However, Anthropic argues this designation lacks factual basis, pointing to its security clearances and past government praise. Two senior Pentagon officials reportedly stated there was "no evidence of supply-chain risk" and the action was "ideologically driven."
Anthropic is pursuing five legal claims, asserting violations of the Administrative Procedure Act, First Amendment, Fifth Amendment, and presidential statutory authority.