Technology and Science30 July, 20263 min read
Judge Rita F. Lin Questions Pentagon Evidence Labeling Anthropic a Supply Chain Risk
This article is an automated aggregation of coverage from news outlets. Our editorials offer in-depth analysis curated by the NewsCord team.
Judge Lin questions DoD's evidence justifying Anthropic's supply-chain risk designation. Designation followed Anthropic's refusal to permit government use of its models.
Lin challenges supply-chain label
A federal judge in San Francisco questioned the Trump administration’s evidence for labeling Anthropic a “supply chain risk” and said she was “troubled” by the idea that a contractor’s public criticism could make it untrustworthy.
U.S. District Judge Rita F. Lin said, “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,” and she added, “I find that position really troubling and at odds with the First Amendment.”
Lin said she had not seen evidence in the record justifying the designation and said, “it appears to me similar to the record I assessed in the preliminary injunction stage,” while also stating, “I don’t see additional evidence really justifying what the government did.”
The dispute centers on the Pentagon’s designation after Anthropic refused to remove restrictions on Claude’s use for lethal autonomous warfare and mass surveillance of Americans, and the case is now before Lin on cross-motions for summary judgment.
Lawfare’s Roger Parloff reported that Lin held the hearing on July 30 in Anthropic PBC v. U.S. Department of War, where the company seeks a permanent injunction against the designation under 10 U.S.C. § 3252.
Government cites distrust, judge skeptical
At the hearing, Department of Justice attorney James Harlow argued that “it was a lack of trust, not retaliation, that led to labeling Anthropic as a national security risk,” and he said AI technology can’t be disassembled and inspected like physical products such as rifles.
Harlow told the court that the government needs to trust AI vendors to be upfront with their technology, and he argued that Anthropic raised concerns after it refused to permit the government to use its technology for “all lawful uses.”
In rebuttal, Michael Mongan of WilmerHale said there was “no evidence Anthropic took steps to interfere with operations” and argued that the government had always known about the company’s two use restrictions against domestic surveillance of Americans and lethal warfare.
Lin pressed the government’s theory that the company could sabotage models, and News of the United States reported she said, “I find that position, if that’s really what the government’s position is, to be really troubling,” while also describing it as “quite extreme.”
The same report said Lin saw no evidence Anthropic could alter its AI models after delivery, flip a kill switch, or seed hidden controls, and it quoted her questioning whether the record showed “any information” that Anthropic posed a threat as of Feb. 27.
What happens next for agencies
Lin’s scrutiny comes as the Pentagon continues to wind down its use of Anthropic products, with Harlow telling the court that the process is scheduled to be completed by Sept. 3 and that certain agencies plan to transition to other AI providers.
News of the United States reported that Lin said the Justice Department declined to say in open court whether any agencies had increased their use of Claude for national security work since the injunction, and she said that conduct could shed light on the government’s original intent.
The Courthouse News report said Lin had issued a preliminary injunction in March prohibiting the government from implementing, applying, or enforcing directives designating Anthropic as a supply chain risk and prohibiting military contractors from doing business with the company.
TechCrunch said Lin agreed the administration still lacked enough evidence, and it quoted her saying she saw no proof Anthropic could alter a delivered model or “flip some kind of kill switch.”
With the next hearing and final decision pending, the dispute remains tied to Anthropic’s refusal to allow mass surveillance of U.S. citizens and lethal autonomous warfare, and to the government’s argument that undisclosed guardrails could risk warfighters and civilians.