A federal judge in California has ruled that the Trump administration’s classification of Anthropic as a “supply chain risk” was unlawful. This decision was made by U.S. District Judge Rita Lin, who stated that the Defense Secretary’s actions amounted to “unlawful retaliation” against the AI company in violation of the First Amendment, and characterized the decision as “arbitrary and capricious.” Furthermore, Judge Lin noted that Anthropic was denied due process, as stipulated by the Fifth Amendment.
The controversy arose earlier this year when Secretary Pete Hegseth and then-President Donald Trump labeled Anthropic a risk to national security, instructing all federal agencies, including those outside of defense, to cease collaboration with the firm.
The tensions were rooted in Anthropic’s refusal to allow the Pentagon to fully utilize its AI models for autonomous weapons and mass surveillance without stringent safety guardrails. In response, the Pentagon denied any intentions to exploit these models for unlawful purposes and argued that Anthropic was trying to control how the military utilized technologies that it had purchased.
In her ruling, Judge Lin emphasized that the government seemed motivated by a desire to publicly reprimand Anthropic for its critique of governmental policies. She highlighted inconsistencies in the government’s approach, including a proposal to invoke the Defense Production Act for Anthropic, which suggests the company is vital to national security rather than a threat. Despite this designation, the Department of Defense continued its pursuit of a contract with the company and collaborated with its cybersecurity model, Mythos.
Lin further asserted that it is evident Anthropic does not retain any backdoor access to its technology once provided to the Department of Defense. “While the Department of War is undoubtedly free to choose its AI vendor, the sweeping measures enforced against Anthropic were both illegal and without merit,” she wrote. She added, “The invocation of national security is not a free pass to penalize and retaliate against critics of the government.”
An Anthropic spokesperson expressed gratitude for the ruling, stating, “We welcome the court’s decision that the supply chain risk designation was unlawful. Our focus remains on working productively with the government to utilize AI for national security, benefiting all Americans.”
In March, Anthropic filed two separate lawsuits against the Department of Defense in California and Washington, D.C., with the latter case still ongoing.
TechCrunch has reached out to the Department of Defense for a statement.
Editor’s Take
This ruling is significant for the AI sector as it emphasizes the importance of due process and free speech, particularly for tech companies that engage with government entities. The court’s decision may lead to a more open dialogue about the ethical use of AI in national security while ensuring that companies can maintain integrity and control over their technologies. It sets a precedent for future engagements between AI firms and government, fostering accountability in regulatory practices.
Source: techcrunch.com
