OpenAI will integrate ChatGPT into GenAI.mil, the Pentagon’s generative‑AI platform, in early July, a company representative confirmed on Tuesday.
The announcement was made by Mohammed Husain, the Chief Digital and Artificial Intelligence Office’s strategic delivery lead for cyber, during the Defense One Tech Summit in Arlington, Virginia.
“We’re going live very soon and look forward to making a broader announcement in early July,” Husain said.
This rollout will give more than three million defense personnel access to ChatGPT, with the service certified for handling controlled unclassified information at Impact Level 5. The Pentagon introduced GenAI.mil in December, initially planning to incorporate Google’s Gemini for Government before later announcing the addition of AI models from OpenAI and xAI. By late April, senior defense officials reported that over 1.3 million users were regularly accessing the platform, which has facilitated the creation of more than 100,000 AI agents.
Federal agencies have been using ChatGPT since at least January 2025. In August, OpenAI offered its model at a discount through a OneGov agreement with the General Services Administration. Earlier this month, OpenAI’s latest version, ChatGPT 5.4, became available to the federal workforce via Amazon Bedrock and GovCloud.
Husain noted that users are likely to demand more tokens—units of data processed by AI—and models that use them more efficiently.
“These models consume a large number of tokens, and completing high‑value work often requires even more,” he explained. “Consequently, ‘token efficiency’ will become a central topic.”
He clarified that token efficiency concerns cost per completed task rather than processing speed. The June release of OpenAI’s GPT‑5.5, GPT‑5.4, and Codex on Amazon Bedrock is expected to support the deployment of more capable, token‑intensive models.
“Deploying these models will make them more intelligent and token‑hungry, so cost efficiency will become a critical issue,” Husain said.
Government agencies are also seeking greater computing power for multicloud and on‑premise environments, a need that providers such as AWS are rushing to meet.