SAN FRANCISCO — A federal judge has ruled that the Pentagon acted unlawfully when it designated the artificial intelligence company Anthropic as a national security supply chain risk, finding that the government retaliated against the company for criticizing its approach to military uses of AI.
U.S. District Judge Rita F. Lin ruled that the actions against Anthropic violated the company’s First Amendment rights and denied it the process required under the Fifth Amendment. She also found that Defense Secretary Pete Hegseth’s decision to classify the company as a supply-chain risk violated federal law and was arbitrary and capricious.
The ruling represents a significant victory for Anthropic in a dispute that began after negotiations between the company and the Pentagon broke down earlier this year.
Also Read: Judge Questions Trump Birthright Citizenship Order
Anthropic, which develops the Claude family of AI models, had been working with U.S. national security agencies. The company said its technology had been deployed for uses including intelligence analysis, operational planning and cybersecurity.
The disagreement centered on restrictions Anthropic wanted to maintain on two uses of its technology: mass domestic surveillance and fully autonomous weapons.
Anthropic said it supported the use of artificial intelligence for national security but argued that current AI systems were not reliable enough to operate fully autonomous weapons without appropriate safeguards. It also objected to the use of its models for mass surveillance of people in the United States.
Pentagon officials had sought agreements that would permit AI companies’ technology to be used for lawful military purposes without those restrictions.
After the two sides failed to reach an agreement, Hegseth moved in February to designate Anthropic as a supply-chain risk. The designation threatened the company’s ability to work with the federal government and companies doing business with the military. President Donald Trump’s administration also directed federal agencies to stop using Anthropic technology.
Anthropic subsequently sued the government.
In her ruling, Lin concluded that the government had failed to establish an adequate national security justification for the measures against the company.
The judge found that the record instead showed that officials had sought to punish Anthropic after it publicly criticized the government’s position. She concluded that the government’s actions amounted to unlawful retaliation protected by the First Amendment.
The court also found problems with the procedure used to impose the designation.
According to the ruling, Anthropic did not receive the process required before the government took action against it, resulting in a violation of the Fifth Amendment. Lin additionally found that Hegseth’s supply-chain designation did not comply with the statute the Pentagon relied upon.
The dispute has become one of the most prominent examples of growing tensions between the U.S. government and artificial intelligence companies over how advanced AI systems should be used by the military.
AI developers have increasingly competed for federal and defense contracts as the Pentagon expands the use of artificial intelligence across intelligence, logistics, cybersecurity and military planning.
At the same time, the rapid development of increasingly capable AI systems has raised questions about the role private companies should play in setting limits on military applications of their technology.
Anthropic has maintained that the Pentagon, rather than technology companies, should make military decisions, but has argued that certain uses require safeguards because of the limitations of current AI technology.
The company said after the dispute began that it remained willing to work with the U.S. government on national security applications while maintaining its restrictions on autonomous weapons and domestic mass surveillance.
The administration can challenge the latest ruling through the federal appeals process.
