Judge Blocks Pentagon Blacklist of Anthropic as Unlawful

A federal judge has halted the Pentagon's effort to blacklist AI company Anthropic, ruling that the Defense Department crossed into unlawful retaliation by punishing the firm for its public objections to military uses of its technology. The decision from U.S. District Judge Rita Lin in California bars enforcement of the supply chain risk designation that had cut Anthropic off from government contracts and pressured private defense contractors to avoid the company.

The clash began earlier this year when the Trump administration sought to integrate Anthropic's Claude models into classified military systems. Anthropic pushed back, insisting its AI not be deployed for mass domestic surveillance or fully autonomous weapons systems that could fire without human oversight. Defense Secretary Pete Hegseth responded by labeling the company a national security threat, a move that effectively isolated it from Pentagon-related business.

Lin's 59-page order found those steps violated the First Amendment and due process protections. She wrote that the "empty invocation of national security is not a blank check to punish and retaliate against government critics." The ruling described the designation as driven by a desire to make an example of the company rather than any concrete evidence of risk to supply chains.

Anthropic welcomed the outcome and said it remains open to working with the government on national security applications that align with its safety standards. The company had argued from the start that its models simply are not reliable enough for battlefield decisions that bypass human judgment.

The Pentagon's approach drew criticism for treating a private firm's policy disagreement as a security threat. Critics of the move noted that similar designations have historically targeted foreign adversaries or clear espionage risks, not American companies voicing concerns about AI ethics. The ruling leaves the military free to choose other AI providers but prevents it from using blacklisting tactics to sideline dissenters.

This case highlights ongoing tensions between rapid AI adoption by the armed forces and private sector limits on how the technology should be applied. The government maintains it needs unrestricted access to the best available tools for lawful operations, including surveillance and targeting systems. Anthropic and like-minded firms counter that pushing AI into those roles without safeguards risks both technical failures and erosion of accountability.

Judge Lin's order does not force the Pentagon to purchase from Anthropic or halt work with competitors. It simply stops the administration from weaponizing regulatory tools against companies that refuse certain applications. The decision stands as a check on executive branch power when national security claims appear tied more to silencing criticism than to verifiable threats.

The administration is expected to appeal, setting up further court review of how far the military can go in shaping private AI development through contract leverage. For now, the ruling preserves space for companies to set boundaries without facing blanket exclusion from government work.