The Department of Defense (DOD) announced Monday it has added Starshield AI’s Grok for Government and OpenAI’s ChatGPT Mil to its GenAI.mil platform. These artificial intelligence assistants are designated for unclassified military use, expanding the range of commercial AI products available to department personnel.
The GenAI.mil platform, which launched last December, initially featured only Google’s Gemini for Government. According to the Pentagon, the platform currently serves more than 1.7 million users. The new additions follow separate deals reached by Elon Musk’s xAI firm and OpenAI earlier this year to provide models for classified settings.
Pentagon officials stated that Grok for Government will offer features such as deep-thinking inference, adaptive reasoning modes, and customizable workspaces. The agency reported that ChatGPT Mil is intended to support document-heavy tasks, including planning, policy, logistics, and administration, for its more than 3 million personnel.
The expansion comes as the department phases out its use of Anthropic’s platforms. The Pentagon previously designated Anthropic as a national security supply chain risk after the company requested guarantees that its Claude model would not be used for autonomous weapons or mass surveillance. Although a federal judge recently struck down that designation as "illegal and baseless," Chief Technology Officer Emil Michael stated the Pentagon plans to complete the removal of Anthropic’s platforms by the end of September.
The financial scale of these AI integrations is partially reflected in existing government contracts. Under a deal with the General Services Administration (GSA), the agency that manages government procurement, federal agencies can purchase xAI’s Grok models at a rate of 42 cents through March 2027. This pricing structure suggests a move toward low-cost, high-volume deployment of AI across both military and civil agencies, rather than restricted, high-cost experimental use.
The transition also sets a precedent for how the U.S. military manages private sector partnerships and ethical constraints. The department’s decision to move forward with xAI and OpenAI while removing Anthropic following a dispute over autonomous weapon restrictions highlights a shift in vendor requirements. The Pentagon will continue the removal of Anthropic platforms through the end of September, while the new Grok and ChatGPT tools are available for immediate use to support unclassified military operations and administration.