U.S. Military & AI

Reports say U.S. forces used Anthropic’s Claude AI during the classified raid that captured Nicolás Maduro, with Axios and other outlets reporting the model was used in the active operation and is reportedly the only major frontier model currently available on classified Pentagon networks, possibly accessed via Anthropic’s partnership with Palantir. Anthropic says it cannot comment on specific operations but insists Claude use must follow its policies banning violence, weapons development and surveillance, even as it negotiates terms with the Pentagon and some officials have raised concerns that could jeopardize a potential contract worth up to $200 million.