Pentagon deploys ChatGPT Mil and Grok for Government for 3 million DoD staff

Full text

Pentagon adopts two new AI models for use by DoD as part of plans to adopt AI technologies

ChatGPT Mil fills the role of a day to day AI assistant, while Grok for Government provides military users with "immediate productivity gains"

Both systems will be offered through the Pentagon's GenAI.mil platform

In order to carry out the Department of Defense’s mission of securing the United States, the Pentagon has rolled out two AI models for its staff that are specifically adapted to “warfighter needs.”

The models are custom variants of OpenAI’s ChatGPT and xAI’s Grok. The platforms are now available for use by the DoD’s 3 million civilian and military staff.

The AI models, called ChatGPT Mil and Grok for Government, are an extension of the Pentagon’s mission to improve productivity and efficiency without having the data input into AI systems analyzed by the same systems as their consumer counterparts.

Pentagon continues mission to procure secure AI

The Pentagon is already operating a ‘secured’ version of Google Gemini as part of the GenAI.mil platform. Of the 3 million staff, 1.7 million are actively using GenAI.mil , with that number likely to increase as more AI models are added.

The Pentagon’s press release states that ChatGPT Mil will be used in much the same way as anyone else would use the consumer version. To that end, it will offer support for routine work and document-heavy tasks, alongside other uses and integrations with chats and files.

As for the secured Grok variant, the press release states, “Starshield AI’s Grok for Government will provide the warfighter with immediate productivity gains, stronger knowledge continuity and more secure and efficient collaboration.”

This allows staff “to execute missions faster and with greater precision across numerous operational contexts, ranging from market research analysis for acquisition professionals to supply-chain management for logisticians.”

The Pentagon’s adoption of AI technology hasn’t been without issues however. Anthropic was designated a supply chain risk for refusing to allow the Pentagon to use Claude for autonomous weapon systems and mass domestic surveillance. That designation was recently overturned .

Other AI providers that work with the Pentagon have taken heed of the example made of Anthropic, with both Google and OpenAI allowing the Pentagon to use their models for “any lawful government purpose” .