The Pentagon’s expansion of GenAI.mil shows generative AI moving beyond military experimentation toward becoming ordinary workplace infrastructure for millions of defense personnel handling everything from logistics and contracting to sensitive unclassified information. (Source: Image by RR)

Pentagon Accelerates Artificial Intelligence Adoption across Daily Operations

The Pentagon is expanding access to generative artificial intelligence across its workforce by adding specialized military versions of ChatGPT and Grok to GenAI.mil, its secure internal AI platform. The system, which originally launched with a government version of Google’s Gemini, already has approximately 1.7 million users and is intended to serve a Department of War workforce of roughly 3 million military and civilian personnel. The expansion, as noted in fortune.com, follows contracts worth up to $200 million each awarded to OpenAI and Elon Musk’s SpaceXAI, alongside contracts with Google and Anthropic, as defense officials push AI deeper into everyday government operations.

ChatGPT Mil will operate inside a protected government environment designed for military requirements, with OpenAI saying user data will remain inside that environment and will not be used to train its models. Both ChatGPT Mil and Grok for Government are approved for Controlled Unclassified Information at Impact Level 5, allowing personnel to use the systems with sensitive government material that is not formally classified. Much of the immediate opportunity is relatively mundane rather than battlefield-oriented: military organizations generate enormous volumes of documents and administrative work involving logistics, contracting, personnel management, acquisition and research that generative AI could potentially accelerate.

The expansion is part of a broader Pentagon strategy to establish an “AI-first” culture in which artificial intelligence becomes integrated into everyday military workflows rather than treated as a specialized technology. Defense leadership has urged personnel to incorporate GenAI.mil into their daily routines, arguing that widespread AI adoption is necessary to preserve U.S. military competitiveness. The rollout has not always been smooth: when the platform originally appeared on government computers with an “I want YOU to use AI” message, some personnel reportedly suspected that the unexpected pop-up itself might be a cyberattack.

One notable absence is Anthropic’s Claude, following a politically charged dispute between the company and the Pentagon over proposed restrictions involving autonomous weapons and surveillance. The government subsequently moved to designate Anthropic a supply-chain risk and phase out its technology, although a federal judge recently ruled that the Pentagon acted unlawfully in making that designation. The episode highlights a broader issue accompanying military AI adoption: as frontier models become increasingly embedded in defense infrastructure, disagreements over acceptable uses, safety restrictions and government authority are becoming strategic questions rather than simply commercial contract negotiations.

read more at fortune.com