Google expands Pentagon access to Gemini after Anthropic pullback
Google expands Pentagon access to Gemini after Anthropic pullback
Google signs a new DoD deal to broaden access to Gemini, following Anthropic's refusal to support domestic surveillance and autonomous weapons.
Google has broadened the Defense Department's access to its Gemini AI, marking another step in the Pentagon's widening use of commercial AI tools. The move follows Anthropic's decision not to allow the DoD to use its AI for domestic mass surveillance and autonomous weapons, a stance that has reshaped the AI vendor landscape and prompted the department to diversify its options.
Under the new agreement, DoD users will gain expanded access to Gemini capabilities across both classified and unclassified networks, enabling faster data processing, improved decision support, and enhanced analysis for military missions. The expansion is designed to bolster resilience and speed in defense workflows while keeping governance and safety considerations in focus.
The Pentagon has already signed AI deals with OpenAI and Elon Musk’s xAI to run tools on classified networks, a sign of how multiple vendors are being brought into critical defense applications. Anthropic’s refusal has intensified the push to avoid overreliance on a single model, a point echoed by senior defense officials who stress the value of having multiple providers available.
A senior Pentagon AI official has warned that relying on one model is never ideal, and broadening access to several providers strengthens resilience and oversight. Analysts say the latest move signals a broader push to integrate commercial AI into defense operations, while policymakers weigh questions about privacy, ethics, and governance as the Department scales up its use of advanced AI tools. The implications for contractors and policy directions could be significant as the DoD seeks speed without sacrificing safeguards.
