In a pivotal ruling for the future of artificial intelligence governance and corporate freedom of speech, a US federal judge has blocked the Pentagon’s order branding AI company Anthropic as a national security risk. The decision comes after months of conflict between the Department of War and the AI developer over the military use of Anthropic’s language model, Claude. The court found that US officials likely violated the law and retaliated against the company for publicly voicing concerns about the potential applications of its technology.

The dispute traces back to the Pentagon’s efforts to expand its use of advanced AI systems for defense purposes. According to court filings, the Department of War instructed government contractors to cease using Anthropic’s Claude system after the company refused to allow its technology to be deployed for all “lawful” military purposes. Anthropic, a leading developer of large language models, had repeatedly warned that its technology could be misused for domestic mass surveillance or fully autonomous weapons systems if left unchecked.

US District Judge Rita Lin, presiding over the case, described the Pentagon’s designation as “classic” First Amendment retaliation. In her March 26 ruling, she blocked not only the security risk designation but also the Department’s order to terminate all government contracts with Anthropic. Judge Lin’s decision emphasized that the move violated constitutional protections and overstepped legal boundaries, stating, “Nothing in the governing statute supports the Orwellian notion that an American company may be branded a potential adversary… for expressing disagreement with the government.” She further noted that such designations are typically reserved for “foreign intelligence agencies, terrorists, and other hostile actors.”

The ruling marks a rare judicial rebuke of the US military in the context of AI and corporate governance. Anthropic filed its lawsuit against the administration of US President Donald Trump on March 23,........

© Blitz