The Pentagon will reportedly adopt Palantir as its main AI system after a fallout with Anthropic.
The firm’s Maven model is already deeply embedded in US defense: It gathers battlefield data and identifies targets. The new move designates it the “program of record,” which streamlines its adoption across the military, Reuters reported.
The decision follows a high-profile dispute between the Pentagon and Anthropic, after the latter refused to permit its Claude model to be used for mass surveillance or fully autonomous weapons; the government has moved to render Anthropic a “supply chain risk,” meaning no other defense suppliers can use it. A complicating factor is that Maven itself uses Claude to analyze the data it collates.