US Media: Pentagon Threatens to Cut Off Collaboration with Anthropic's AI

On February 15, a senior U.S. government official told Axios that the Pentagon is weighing ending its collaboration with AI firm Anthropic, as the company is imposing restrictions on the military’s use of its models. The Pentagon has been pushing the four leading AI labs to allow military use of their tools for “all lawful purposes” — including sensitive areas like weapon development, intelligence collection, and battlefield operations. Anthropic has rejected these terms, and after months of tense negotiations, the Pentagon has grown frustrated. The company insists two areas must remain off-limits: mass surveillance of the American public and fully autonomous weapon systems. The official noted there is significant gray area around which scenarios fall into these categories. Additionally, the Pentagon finds it unworkable to negotiate with Anthropic for every specific use case, or if Anthropic’s Claude tool unexpectedly blocks certain applications. “Everything is on the table,” the official said — including temporarily scaling back collaboration or fully terminating the partnership. “But if we decide this is the right move, we need to find an appropriate replacement for them.” (Source: Golden Finance)