Judge likely to rid Anthropic of Pentagon’s ‘supply chain risk’ label

The judge previously found the government likely violated the law and temporarily blocked the designation.SAN FRANCISCO (CN) — A federal judge indicated Thursday she is likely to permanently block the government from labeling artificial intelligence developer Anthropic as a “supply chain risk,” saying she was “troubled” by the idea that a company’s public criticism of the government makes it untrustworthy.“The government’s position is that if a government contractor goes out and publicly criticizes the administration, the government can turn around and say ‘I don’t trust you’ and retaliate against the contractor. I find that position really troubling and at odds with the First Amendment,” U.S. District Judge Rita F. Lin said. “I would like to understand better if that is really what the government’s position is because that is really quite extreme.”Lin said from the bench she had not seen evidence in the record justifying the government’s designation of Anthropic as a supply chain risk, nor had she seen evidence to support the government’s arguments that the tech company attempted or will attempt to alter or control its products supplied to the government for defensive purposes.“Having reviewed the record the parties have submitted, it appears to me similar to the record I assessed in the preliminary injunction stage,” the Joe Biden appointee said. “I don’t see additional evidence really justifying what the government did; if anything, it seems like the record has gotten worse for the government.”On Feb. 27, Trump ordered government agencies to stop using Anthropic’s technology, calling it a “radical left, woke company” on social media.The same day, the Pentagon announced it was designating Anthropic a supply chain risk to national security, with Secretary of Defense Pete Hegseth posting on X: “Effective immediately, no contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic … This decision is final."According to Anthropic, Hegseth specifically took issue with the company’s refusal to remove two narrow contractual restrictions on Claude’s use for lethal autonomous warfare and the mass surveillance of Americans. It has no confidence Claude would be reliable or safe if it supported such warfare, the company said.In March, Lin issued a preliminary injunction prohibiting the government from implementing, applying, or enforcing the directives designating Anthropic as a supply chain risk and prohibiting military contractors from doing business with the company, as well as a March 3 letter from Hegseth formalizing the designation.At Thursday’s hearing on opposing motions for summary judgment, Department of Justice attorney James Harlow argued it was a lack of trust, not retaliation, that led to labeling Anthropic as a national security risk.Harlow said AI technology, unlike other physical products the military procures, such as rifles, can’t be disassembled and inspected, meaning the government needs to trust AI vendors to be upfront with their technology.“Nothing prevents Anthropic in the future from developing new policy positions that can be baked into models for additional guardrails that may or may not be disclosed to the Department of War,” he said.Specifically, Harlow said Anthropic raised concerns after it refused to permit the government to use its technology for “all lawful uses,” and a high-level Anthropic executive questioned the use of the technology during an active overseas military operation.When asked what the question was, Harlow declined to specify, but said it concerned “whether it was right for Anthropic technology to support this active military action.”In rebuttal, Michael Mongan of WilmerHale, representing Anthropic, said the executive at issue raised the question to a third party, and there was “no evidence Anthropic took steps to interfere with operations” of the government.Mongan said the government had always known the company’s two use restrictions against domestic surveillance of Americans and lethal warfare, adding it was “hard to swallow” how a “contractual impasse” was all of a sudden elevated to supply chain risk on Feb. 27.“There is nothing in this record that remotely substantiates the suggestion that Anthropic is a national security risk,” he said.Following Lin’s preliminary injunction in March, most government agencies continue to use Anthropic, with two agencies pausing their use and others negotiating new contracts, according to Mongan.Harlow said the Department of War is in the process of winding down use of Anthropic products, which is scheduled to be completed by Sept. 3. Certain agencies plan to transition from Anthropic to other AI providers, while others have reported not stopping use of Anthropic at all.Harlow objected to a question about agencies expanding use of Anthropic in national security work, saying the question could set a “dangerous precedent calling for government-wide information about national security tools.”In response, Lin said it would be “inconsistent” for the government to publicly state it believes Anthropic to be a national security threat yet continue to expand use of its AI model.Lin did not indicate when she would release a ruling.A representative for Anthropic declined to comment.A representative for the Department of Justice did not immediately respond to a request for comment.Subscribe to our free newslettersOur weekly newsletter Closing Arguments offers the latest about ongoing trials, major litigation and rulings in courthouses around the U.S. and the world, while the monthly Under the Lights dishes the legal dirt from Hollywood, sports, Big Tech and the arts.Additional Reads