Federal court in the USA, Pentagon’s Anthropic supply chain risk He found the decision to blacklist him unlawful. Judge Rita F. Lin ruled Aug. 27 that the government’s steps against Anthropic constituted unlawful retaliation for the company’s criticism. However, the court also concluded that Anthropic did not have sufficient opportunity to object before the decision. The dispute over Claude’s military use, which began months ago, resulted in a major court decision against the Pentagon.
Dispute between Anthropic and Pentagon, Claude’s military terms of use It grew as the parties could not agree on the issue. Anthropic announced in February that it supported the US military’s use of Claude for legal purposes, but would not allow two uses. The company these Mass surveillance of US citizens And fully autonomous weapons He defined it as: Anthropic argued that today’s advanced artificial intelligence models have not reached the level of being able to reliably manage fully autonomous weapons. On the other hand, the company means completely removing human intervention with fully autonomous weapons and automating the decisions to select and attack the target. Anthropic’s official statement also states that these two exceptions have not blocked any US government missions to date.
The Pentagon, on the other hand, argued that the government, not private companies, should decide which artificial intelligence is appropriate to use in military activities. Defense Secretary Pete Hegseth, after talks failed On February 27, 2026 Instructed to identify Anthropic as a supply chain risk. Then, the Pentagon officially announced this definition in a letter sent to Anthropic on March 4. In response, Anthropic argued that the law was intended to protect the government’s supply chain and could not be used to penalize a supplier. The company also clarified that the limitation could only cover the use of Claude in Pentagon contracts and not other commercial relationships.
While the Trump administration wanted federal agencies to stop using Anthropic products, Hegseth also tried to limit military contractors’ relationships with the company. Thereupon Anthropic, on March 9 He appealed to federal court in California and launched a legal fight against the Pentagon decision. The company also appealed the second decision, which it received under a different federal law, to the federal appeals court in Washington. However, the California lawsuit did not focus solely on which AI vendor the Pentagon would work with. The court also examined the legal authority the government used to limit Anthropic’s relationships with other entities and military contractors.
The court found Pentagon’s Anthropic decision unlawful
Judge Lin cited the Pentagon’s use in his earlier provisional injunction. supply chain risk raised serious legal objections to its definition. According to the court, the relevant law targets security threats, such as when an adversary introduces harmful functions into military systems or sabotages them. On the other hand, Anthropic’s disagreement with the Pentagon over contract terms is not included in this scope, according to the court’s evaluation. Additionally, the judge noted that the Pentagon had not convincingly explained why more limited options would be inadequate. The court therefore concluded that the legal tool used by the government was not appropriate for its contract dispute with Anthropic.
In the final decision dated 27 August 2026 Lin determined that Anthropic’s public criticism of the government was protected under the First Amendment of the U.S. Constitution. The court found the government’s subsequent steps against Anthropic to be unlawful retaliation for these statements. Additionally, Lin ruled that the government’s failure to provide Anthropic an opportunity to respond to the charges before the ruling violated its Fifth Amendment right to due process. However, the court also found that Hegseth’s supply chain risk decision did not comply with applicable federal law. The decision thus covered both constitutional rights and how the Pentagon used the law in question.
The court’s decision condemns the Pentagon claude It does not force you to use or enter into a new contract with Anthropic. The federal court has previously made clear that the Department of Defense has the right to choose the artificial intelligence provider it wants. On the other hand, the government went beyond choosing not to work with a specific company and targeted Anthropic’s other business relationships. The court blocked federal agencies from enforcing the Trump administration’s order to halt the use of Anthropic and struck down the Pentagon’s designation as a supply chain risk. While the U.S. government may appeal the decision, Anthropic’s second legal action, dealing with a different federal law, continues in Washington.
According to Anthropic’s own statements, Claude has been working in the secret networks of the US government since June 2024. The company states that its models are used in military fields such as intelligence analysis, modeling and simulation, operation planning and cyber operations. In contrast, Anthropic continues to maintain its two exceptions for mass internal surveillance and fully autonomous weapons. The company also announced in March that if the Pentagon wanted to make the transition, it would continue to offer its models at a low price until the transition was completed.
Join Channel