Court finds unlawful retaliation
SAN FRANCISCO — A federal judge ruled Thursday night that the Pentagon unlawfully retaliated against artificial-intelligence company Anthropic after it publicly opposed unrestricted military use of its Claude models.
U.S. District Judge Rita F. Lin found that the challenged actions violated Anthropic’s First Amendment rights and that the company was denied the process required by the Fifth Amendment before the government imposed sweeping penalties. The 59-page decision also concluded that Defense Secretary Pete Hegseth’s supply-chain-risk designation violated the governing statute and was arbitrary and capricious.
The ruling resolves key claims on the merits rather than merely predicting how the case might end. Lin had issued preliminary protection earlier in the litigation, but Thursday’s summary-judgment decision represents a substantially stronger legal defeat for the government.
Read the source: U.S. District Court: August 27 summary-judgment order ↗
Dispute centered on surveillance and autonomous weapons
The confrontation began after Anthropic refused to remove restrictions that it said were intended to prevent Claude from being used for mass domestic surveillance or fully autonomous weapons. The Pentagon sought permission to use the company’s technology for any lawful military purpose.
President Donald Trump and Hegseth subsequently accused the company of endangering national security. The government designated Anthropic a supply-chain risk, moved to end federal use of its products and directed restrictions affecting defense contractors that worked with the company.
Anthropic argued that the government was punishing protected speech and trying to force a private company to abandon publicly stated safety limits. The administration maintained that national-security officials needed broad discretion over vendors supporting military operations.
Read the source: Associated Press: Judge rules for Anthropic in Pentagon dispute ↗
Judge says record did not support security rationale
Lin found that the government’s stated national-security explanation did not match the administrative record. The opinion says the measures were driven by a desire to make a public example of Anthropic for criticizing the administration, not by evidence that the company would sabotage its technology.
The court emphasized that the Pentagon remains free to choose another vendor and is not required to purchase Anthropic products. The legal problem, Lin concluded, was the government-wide punishment and secondary restrictions imposed in response to the company’s public position.
The decision was not a complete victory for Anthropic. Lin rejected its separation-of-powers claim and entered judgment for the government on claims involving agencies that had not participated in the challenged conduct.
Read the source: Reuters: Judge rules Pentagon blacklisting of Anthropic unlawful ↗
Exact final relief and appeal remain ahead
The court said a separate order will specify the final relief. Lin denied the government’s request to delay the anticipated permanent injunction for seven days, but Thursday’s opinion did not itself describe every operational step the Pentagon and other agencies must take.
The Associated Press reported that the government is expected to continue fighting the decision. A separate, narrower Anthropic case involving another Pentagon rule remains pending in the federal appeals court in Washington, D.C.
The case could shape how far a president and defense officials may go when a technology contractor publicly challenges military policy. It also leaves unresolved the broader policy question of who should set enforceable limits on artificial intelligence used for surveillance and autonomous weapons.
Read the source: Wikimedia Commons: Public-domain Pentagon photograph ↗
Sources and further reading
U.S. District Court: August 27 summary-judgment order ↗
Associated Press: Judge rules for Anthropic in Pentagon dispute ↗
Reuters: Judge rules Pentagon blacklisting of Anthropic unlawful ↗