The Pentagon Just Got Its Own ChatGPT for 3 Million Troops

The Pentagon now has its own private versions of ChatGPT and Grok, two of the most popular AI chatbots. They're part of a secure military portal called GenAI.mil, which already had Google's Gemini. The goal is simple: let 3 million military and civilian defense workers use powerful AI without risking sensitive government data on public apps. Unlike consumer chatbots, these military versions don't collect or sell your data. And they're already a hit — more than 1.7 million people have signed up.

What do people actually use them for? Mostly boring but crucial work: writing reports, handling logistics, planning budgets, and reviewing policies. ChatGPT Mil looks and works just like the regular ChatGPT people use at home, but it's been cleaned up for government use. Think of it as a work assistant for soldiers who spend their days buried in paperwork.

The bigger drama is around Grok. The Pentagon described it in much more warlike terms, saying it would help troops "execute missions faster and with greater precision" — from supply-chain management to market research. That's why the military is also using Grok, built by Elon Musk's xAI and accessed through SpaceX's secure satellite network. It's a reminder that AI isn't just for writing emails; it's becoming part of how wars are planned and fought.

And there's a notable player missing: Anthropic, the company behind Claude. It refused to give the Pentagon unrestricted access, insisting on safety guardrails. The Trump administration then labeled Anthropic a security risk, a decision the company is fighting in court. So this isn't just about cool tech. It's about who controls powerful AI and how much safety we're willing to trade for military advantage — especially when taxpayers are footing the bill.