The Pentagon is rapidly expanding its use of generative AI, and this time the rollout puts two of the biggest names in the industry directly in front of the US military workforce.
On August 31, 2026, the Pentagon’s GenAI.mil platform added two new AI assistants, ChatGPT Mil from OpenAI for Government and Grok for Government from SpaceX’s Starshield AI. The move gives nearly 3 million civilian and military personnel access to frontier AI tools for unclassified work, including document-heavy administrative and staff tasks, according to TechCrunch and the Department of War.
READ: OpenAI files a motion to dismiss Apple lawsuit (August 6, 2026)
The scale of the rollout is significant. GenAI.mil was already being used by about 1.5 million personnel as of June, according to the Pentagon’s chief technology officer. Five of the six military branches had adopted the platform as their default AI tool, while users had already created roughly 100,000 AI agents.
The latest additions could make the U.S. military one of the largest single enterprise users of frontier AI models. What makes the development particularly notable is how quickly it has happened. GenAI.mil has expanded to this scale in less than a year, despite remaining relatively unknown outside government and defense circles.
The Pentagon’s use of these tools is focused on unclassified military operations rather than weapons systems. Personnel can use the assistants for tasks such as summarizing doctrine, preparing routine documents and preserving institutional knowledge as service members move between assignments. The Department of War said Grok for Government is intended to improve productivity while helping warfighters maintain access to organizational knowledge.
The companies are also approaching the government market through dedicated channels rather than simply opening access to their consumer products. OpenAI is providing ChatGPT Mil through its OpenAI for Government business, while Grok is being delivered through Starshield AI, SpaceX’s government-focused AI operation.
The Pentagon’s rollout highlights a broader shift in how major AI companies are approaching government customers. Instead of treating public-sector agencies as another enterprise account, companies are increasingly developing specialized versions of their models designed around government requirements, security controls and deployment environments.
READ: OpenAI to pay $3.2 million to settle DOJ visa hiring discrimination case (August 5, 2026)
There is also a bigger lesson in the 100,000 AI agents already created on GenAI.mil. As organizations give employees the ability to build their own AI agents, managing those systems could become just as important as the underlying models. Permissions, oversight, security, auditing and deciding which agents should remain active could quickly become major challenges for large organizations.
The speed of the Pentagon’s AI expansion also stands out against the slower adoption seen in parts of the private sector. Companies in industries such as insurance and compliance are still working through questions around risk, accountability and what happens when AI agents make costly mistakes.
For the Pentagon, however, generative AI is moving from experimentation to everyday workplace infrastructure. With ChatGPT and Grok now joining the platform, the U.S. military is signaling that frontier AI is no longer being treated simply as an emerging technology. It is becoming part of how the world’s largest military organization handles its day-to-day work.