1 September 2026

Pentagon deploys custom ChatGPT and Grok for 3 million military and civilian staff

First reported

TechCrunch ran this on 31 August, a day before the next source picked it up.

The Department of Defense launched ChatGPT Mil and Grok for Government, tailored versions of OpenAI's and xAI's chatbots, through GenAI.mil, a secure portal that prevents sensitive military data from entering consumer AI systems.
1.7 million of the Pentagon's 3 million personnel are already using GenAI.mil, which also includes a secured Google Gemini variant, with adoption expected to grow as more models are added.
ChatGPT Mil handles routine unclassified work like document processing and administrative tasks, while Grok for Government is positioned for broader military applications including mission planning, supply chain management, and faster decision-making across operations.
Anthropic's Claude model is absent from the platform after the company refused to allow Pentagon use of its AI for autonomous weapons and mass surveillance, receiving a supply chain risk designation that it is fighting in court.
OpenAI and Google agreed to let the Pentagon use their models for any lawful government purpose, contrasting with Anthropic's conditional approach and reflecting how military partnerships shape tech companies' safety policies.