DoD Broadens Secure AI Access with OpenAI and xAI Models

The U.S. Department of Defense has expanded its secure GenAI.mil platform to include customized versions of OpenAI’s ChatGPT and xAI’s Grok, offering its civilian and military workforce new generative‑AI tools under strict government‑grade security. The additions, named ChatGPT Mil and Grok for Government, are available through the centralized portal that was originally launched with Google’s Gemini. According to the DoD, more than 1.7 million personnel have already used GenAI.mil, representing a significant portion of the roughly 3 million civilian and military staff in the department.

The initiative is part of Washington’s effort to integrate cutting‑edge AI into everyday defense operations while preventing sensitive data from being exposed to consumer‑grade services. The military‑specific versions operate under tailored security arrangements and do not collect data in the same way that consumer AI platforms do.

ChatGPT Mil is positioned not as a battlefield system but as a versatile knowledge‑work assistant, capable of handling conversations, file management, project support, and custom GPTs. The DoD envisions applications across administration, logistics, planning, and policy work, especially where large volumes of documents must be processed.

The DoD’s partnership strategy extends beyond OpenAI and xAI. Amazon Web Services, Microsoft, Nvidia, and Reflection AI have also secured agreements or collaborations to bring AI capabilities to the department.

Anthropic’s Claude was omitted from the platform after a dispute with the Trump administration over the company’s willingness to grant unrestricted access and maintain safety restrictions. Anthropic has since challenged the supply‑chain risk designation in court.

Overall, the DoD’s broadened AI portfolio underscores its commitment to embedding generative AI into routine defense workflows while maintaining stringent data‑protection standards.