The Pentagon Gets Military Versions of ChatGPT and Grok

The U.S. Department of Defense has rolled out dedicated versions of OpenAI's ChatGPT and xAI's Grok, opening up access to generative AI for three million civilian employees and service members. The tools, dubbed ChatGPT Mil and Grok for Government, have joined the centralized, secure GenAI.mil portal, which went live last year and initially offered only Google Gemini.

The portal's core idea is to give the department's staff access to cutting-edge commercial models without funneling sensitive government data through ordinary consumer channels. The military versions, in particular, are exempt from the data collection that is hard to avoid in mass-market products. Judging by the Pentagon's own figures, the venture is popular: of the three million people, more than 1.7 million unique users have already signed on to the service.

A Different Tone for Different Models

ChatGPT Mil grew out of the OpenAI for Government program and promises an experience familiar to ordinary users: chat, working with files, projects and custom GPTs. It is positioned modestly — as an assistant for routine, unclassified work: administration, logistics, planning and document preparation.

Grok, by contrast, sounds far more martial in the Pentagon's telling. According to the department, Grok for Government — from Starshield AI (SpaceX's secure satellite network built on Starlink technology) — will give the “warfighter” a boost in productivity, continuity of knowledge and safer interaction, making it possible to carry out tasks “faster and more accurately,” from market analysis for procurement specialists to supply-chain management.

A Notable Absence

Against this backdrop, the absence of Anthropic's Claude model stands out. The Trump administration has declared the company a “supply-chain risk” — a decision Anthropic is now challenging in court. The root of the conflict is that the lab refused to grant the Pentagon unlimited use of its tools and insisted on certain safety safeguards.

The portal's expansion is part of the Pentagon's broader strategy to speed up its work through AI without compromising security. In addition to OpenAI and xAI, the department has struck agreements with Amazon Web Services, Microsoft, Nvidia and Reflection AI.

Source: TechCrunch