OpenAI deploys a custom ChatGPT on GenAI.mil, the Department of War's enterprise AI platform for 3 million personnel

OpenAI is deploying a custom version of ChatGPT onto GenAI.mil, the U.S. Department of War's secure enterprise AI platform, giving the department's 3 million civilian and military personnel access to a government-approved version of the chatbot for unclassified work.

What's new

In a post announcing "the next phase of our national security work," OpenAI said it is "bringing ChatGPT to GenAI.mil, the Department of War's secure enterprise AI platform used by 3 million civilian and military personnel." The deployment, branded ChatGPT Mil, went live on GenAI.mil the same day, according to a Department of War release, and is accredited for Controlled Unclassified Information at Impact Level 5 (IL5) — the security tier the Pentagon requires for its more sensitive unclassified systems.

OpenAI describes the deployment as a custom version of ChatGPT, "approved for the Department's unclassified work," that "runs in authorized government cloud infrastructure with built-in safety controls and protections for the Department's highly sensitive data." The company lists specific use cases it's targeting: summarizing and analyzing policy and guidance documents, drafting and reviewing procurement and contracting materials, generating internal reports and compliance checklists, and supporting research, planning, mission support, and administrative workflows. The core experience centers on chat, file uploads, project organization, and custom GPTs, with additional features to be added over time.

On data handling, OpenAI says explicitly that data processed on GenAI.mil "remains isolated to the government environment and is not used to train or improve OpenAI's public or commercial models."

Context

This is not OpenAI's first work with the Pentagon: the company points to its existing collaboration with DARPA on the AI Cyber Challenge and a pilot program announced earlier in 2026 with the Department's Chief Digital and Artificial Intelligence Office (CDAO) exploring how frontier AI could change Pentagon operations. GenAI.mil itself is the Department of War's shared generative-AI platform, hosting tools from multiple frontier AI labs rather than being exclusive to any one vendor — OpenAI's post frames this launch as "joining the other frontier AI labs" already on the platform. The Department of War's own release describes today's launch as the adoption phase of an enterprise partnership the two sides established in 2025.

Why it matters

Getting a frontier chatbot accredited to IL5 and deployed to 3 million Department of War personnel is one of the largest single-organization government AI rollouts OpenAI has landed to date, and it puts ChatGPT inside day-to-day defense administrative and planning workflows rather than in a narrow pilot. It also sharpens the competitive picture in government AI: GenAI.mil already hosts tools from rival labs, so OpenAI's push here is as much about becoming the default assistant for a captive, security-cleared user base as it is about the deployment itself. The explicit no-training-on-government-data commitment matters for adoption inside an institution that is unusually sensitive about data exposure, and sets a data-handling bar other labs on the same platform will likely be measured against.