Judge Temporarily Blocks Pentagon’s Ban, Siding with Anthropic

Judge Lin stated that penalizing Anthropic is a clear violation of the First Amendment due to retaliation. Following a protracted dispute with the Pentagon, a judge granted Anthropic a preliminary injunction in its lawsuit to overturn its government blacklisting as the case proceeds. Judge Rita F. Lin from the northern district of California noted the Department of War flagged Anthropic as a threat due to its negative publicity in the press. She likened this to illegal First Amendment retaliation, with her order becoming effective in a week. A final decision remains pending. Anthropic spokesperson Danielle Cohen expressed gratitude for the court’s swift action and optimism about their chances of success. The company emphasizes its focus on collaboration with the government for safe AI. Judge Lin acknowledged the broader debate: Anthropic claims its AI, ‘Claude,’ is unsuitable for autonomous weapons or mass surveillance, whereas the Pentagon insists on their prerogative to determine its AI use. Lin clarified her role is not to decide tech use matters but to ensure legal protocol was followed. Problems originated from Defense Secretary Pete Hegseth’s directive to integrate ‘any lawful use’ language into AI contracts, impacting negotiations over Anthropic’s refusal to allow its tech for mass surveillance or autonomous weapons. Events included heated exchanges, Anthropic labeled a supply chain risk, and ensuing legal action. Anthropic argues its punitive designation contravenes First Amendment rights, seeking its reversal. The rarity of a U.S. company getting such designation caught public attention due to potential punitive repercussions for dissent from government policies. Court documents reveal significant business impacts for Anthropic, with many partners confused or reconsidering collaboration. The case dialogue included Judge Lin questioning the government’s authority over Hegseth’s directives and the risk justification behind Anthropic’s designation. Tensions rose with Hegseth’s ambiguous social media announcement, prohibiting military partners from dealing with Anthropic. Lin prodded for clarity on implications for military contractors, distinguishing between military-purpose services and unrelated sectors like toilet paper supply. She also questioned the Pentagon’s assertions of risk associated with Anthropic potentially disabling its technology, highlighting lack of evidence to suggest Anthropic could unilaterally impact deployed AI systems.