A US appeals court has upheld the Pentagon's decision to exclude artificial intelligence company Anthropic from its supply chain. The ruling follows a dispute over the company's refusal to remove safeguards on its Claude AI model. The company had been used by the US military across various classified systems, including during a January operation that resulted in the capture and removal of Venezuelan leader Nicolas Maduro.
The dispute began in February when Anthropic refused a demand from US Defence Secretary Pete Hegseth to remove restrictions on the use of Claude for fully autonomous lethal weapons and mass domestic surveillance of Americans. The Pentagon subsequently designated Anthropic as a supply-chain risk in March, leading to the cancellation of the company's military contracts and preventing other Pentagon contractors from using Anthropic's technology.
The move represented an unusual use of the Pentagon's procurement authority against a US technology company and triggered a legal dispute over the government's ability to restrict a company due to disagreements over its technology deployment. Following the Pentagon's decision, the US military turned to Anthropic's competitors, including OpenAI, the company behind ChatGPT, which signed an agreement with the Pentagon on the same day as Anthropic's ban.
Anthropic has remained a major player in the rapidly expanding AI industry despite the dispute, valued at nearly $1 trillion and reportedly preparing for a major initial public offering. The company's Chief Executive Officer, Dario Amodei, has repeatedly warned about potential dangers associated with advanced artificial intelligence, contrasting with President Donald Trump's dismissal of these warnings as a hoax.
Under Secretary of Defense Emil Michael welcomed the appeals court ruling, arguing that it reinforced the government's position in the dispute and protected military personnel from private companies inserting their opinions into the chain of command. However, the legal battle is not over, as the appeals court ruling conflicts with another federal court decision concerning Anthropic.
A federal court in California ruled last month that a broader ban imposed by the Trump administration against Anthropic was unlawful, representing a conflicting judicial finding. The appeals court ruling upheld the Pentagon's specific designation of Anthropic under public procurement regulations, which remains in effect. Anthropic disagreed with the ruling, citing the separate California decision.
The company said it remained confident in its position and was considering further legal options. The case highlights growing tensions between technology companies and governments over the acceptable uses of increasingly powerful artificial intelligence systems, particularly in military operations and surveillance. The Pentagon's decision leaves Anthropic facing restrictions on its participation in the department's supply chain.
Key points
- The US appeals court upheld the Pentagon's ban on Anthropic due to disagreements over the company's refusal to remove safeguards on its Claude AI model.
- The dispute highlights tensions between technology companies and governments over the use of artificial intelligence in military operations and surveillance.
- The ruling conflicts with another federal court decision concerning Anthropic, indicating a potential ongoing legal battle.
