The US Department of Defense has partnered with seven AI firms, including OpenAI and Google, to deploy tools in classified networks. The move reduces reliance on single vendors and accelerates AI adoption. Anthropic remains excluded over security concerns, despite internal demand for its models among defence personnel.