The court in the United States supported the inclusion of Anthropic in the blacklist of the Pentagon

A federal appeals court in Washington, D.C., has refused to block a decision by the U.S. Department of Defense to blacklist AI developer Anthropic. Anthropic does not agree with this decision, is confident in the rightness of the company and is considering new legal steps. Reuters reports.

The court's decision was made on a lawsuit filed by Anthropic in March. In it, the company protested the right of the US Department of Defense to blacklist it and recognize “a risk to the supply chain.” The court found that the Pentagon had good reason for such action, given the fact that Anthropic puts in its AI model Claude software code that can limit the functions of the model or block it remotely. This, according to the court, could be regarded by the Pentagon as a risk to military systems, since such software can at any time stop following orders due to the internal ethical attitudes of the developer.

The conflict between the Pentagon and Anthropic began in February 2026. Then the company refused to provide the military with full access to its AI without the standard restrictions that prevent the combat use of neural networks. The Pentagon in response blacklisted Anthropic and banned its units from using its tools. In March, Anthropic filed appeals in California, challenging the Pentagon's actions as a violation of free speech, and in Washington. In August, a California court sided with the company and blocked Anthropic from being blacklisted. Since the two courts have ruled otherwise, the case could go to the U.S. Supreme Court.

Alena Miklashevskaya