On Tuesday afternoon, Anthropic confronted the Department of Defense in a federal court as the AI company sought a temporary halt to the government’s decision barring the US military and contractors from using its technology. The dispute centers on Anthropic’s refusal to let its Claude AI chatbot be employed for domestic mass surveillance and fully autonomous lethal weapons. This controversy intensified after Donald Trump ordered all US government agencies to cease using Anthropic’s tools, a directive the company is also challenging. The AI firm’s representatives and government officials appeared before Judge Rita Lin in a northern California district court, where the judge oversaw the hearing on a temporary injunction. This session marked an early stage in Anthropic’s lawsuit against the defense department, filed earlier this month following Pete Hegseth, the defense secretary, labeling the company a supply chain risk. Anthropic claims this classification will cause irreparable damage and cost it hundreds of millions or more in revenue. The lawsuit and Judge Lin’s ruling are poised to have extensive consequences for both Anthropic and the US government, which has grown heavily reliant on Claude over the past year for various functions, including military operations against Iran. The conflict, notably the defense department’s unprecedented step to brand a US company as a supply chain risk, has also strained the close ties between Silicon Valley and the Trump administration. Judge Lin described the case as a “fascinating public policy debate” but emphasized her responsibility is narrowly focused on determining whether the government’s actions were illegal. She also expressed suspicions about the government’s conduct, suggesting it may have exceeded merely deciding not to collaborate with Anthropic and moved into punitive territory.

“It looks like an attempt to cripple Anthropic, ” Lin remarked. Government attorneys countered that Hegseth’s social media announcement last month, which barred contractors from working with the government if they engaged with Anthropic, was not a legally binding action. They argued that no party would face penalties for ignoring it. However, their position appeared to conflict with Hegseth’s post on X, which stated that any contractor doing business with the military must avoid Anthropic. Judge Lin pressed the government’s lawyer, saying, “You’re standing here saying, ‘We said it, but we didn’t really mean it. ’” She further questioned why Hegseth would make such a claim if it held no legal weight. “I don’t know, ” responded the government’s attorney. Anthropic declined to comment on the litigation. The defense department has stated it does not discuss ongoing legal matters as a policy. Anthropic alleges that designating it a supply chain risk violated its First Amendment rights, contending the move was intended to punish the company for displeasing the president and resisting the department’s request to relax safety guardrails on Claude. “These actions are unprecedented and unlawful. The constitution does not allow the government to wield its enormous power to punish a company for its protected speech, ” Anthropic stated in its filing in California. The company has maintained that its AI model is not sufficiently reliable for mass domestic surveillance or fully autonomous lethal weapon systems. CEO Dario Amodei has voiced apprehension about AI’s potential use in authoritarian ways. Conversely, US defense officials and Trump have characterized Anthropic’s stance as politically biased and a betrayal, with Trump labeling it a “A RADICAL LEFT, WOKE COMPANY” in a post on his social media platform, Truth Social.