Due to lack of evidence, the Justice Department plans to suspend the "supply chain risk" determination of Anthropic. According to people familiar with the matter and public court records, the U.S. government's attempt to label the artificial intelligence startup Anthropic as a "supply chain risk" and comprehensively ban federal agencies from using its technology has once again been frustrated.

Due to lack of evidence, the Justice Department plans to suspend the determination of "supply chain risk" against Anthropic. According to people familiar with the matter and public court records, the US government's attempt to label the artificial intelligence startup Anthropic as a "supply chain risk" and comprehensively prohibit federal agencies from using its technology has once again been frustrated. At a court hearing on Thursday, U.S. District Court Judge Rita Lin made it clear that the Trump administration has not yet provided enough evidence to support the decision to kill. The legal dispute began with stalled contract negotiations between Anthropic and the U.S. Department of Defense (DOD). Anthropic insists that its AI technology is not yet mature and requires clauses to be added to the agreement prohibiting its AI systems from being used for large-scale surveillance of U.S. citizens, or from directly participating in the targeting and firing decisions of lethal weapons. However, the Pentagon has firmly opposed this, emphasizing that private companies should not impose conditions on the use of technology by the military, and emphasizing that the military will only use these tools within "legal" limits. In addition to disagreements over the scope of use, the U.S. government contends that Anthropic's public criticism of the Department of Defense constitutes grounds for a ban. Judge Rita Lin expressed serious concerns about this logic, calling the approach "deeply troubling" and warning that it could set an extremely dangerous precedent whereby the government could retaliate against federal contractors who dissent. Additionally, the Department of Defense has alleged that Anthropic could remotely disable or modify delivered AI models during military operations. However, many industry technical experts have questioned this statement and believe that it lacks factual basis. Judge Rita Lin also agreed during the hearing, saying there is currently no evidence that Anthropic has the ability to modify the delivered model or "flip some kind of remote kill switch." This hearing is the trial part of one of two lawsuits initiated by Anthropic against the Ministry of Defense in March this year. The company aims to revoke the risk labels and bans established against it through legal means. Judge Rita Lin issued a temporary injunction in March this year to prevent the government's ban from taking effect. She is currently considering whether to convert the temporary injunction into a permanent injunction. via cnBeta.COM - Chinese industry information station (author: source: cnBeta.COM)