Pentagon Grants 3 Million Workers Access to ChatGPT and Grok via Secure GenAI.mil Platform

The U.S. Department of War has announced it is granting 3 million military and civilian personnel access to specialized versions of ChatGPT and Grok. The rollout aims to integrate generative AI into daily workflows through the department's bespoke GenAI.mil platform.

Expansion of GenAI.mil

Launched in December with an initial focus on Google’s Gemini, GenAI.mil already counts 1.7 million users. The inclusion of ChatGPT and Grok follows the awarding of defense contracts worth up to $200 million each to OpenAI and Elon Musk’s SpaceXAI. Contracts were also previously awarded to Google and Anthropic.

The military version, dubbed "ChatGPT Mil," is part of the OpenAI for Government initiative. It is designed to handle Controlled Unclassified Information (CUI) at Impact Level 5 (IL5). OpenAI officials stated that the tool will streamline routine tasks in HR, logistics, and contracting, ensuring that data remains within a secure environment and is not used for model training.

Grok and the ‘AI-First’ Mandate

The military version of Grok, provided by Starshield AI, will offer personnel access to deep thinking-inference and various reasoning levels (auto, fast, and expert). This integration is part of a broader strategy led by War Secretary Pete Hegseth to ensure American dominance through an "AI-first" mentality.

“The future of American warfare is here, and it’s spelled AI,” Hegseth declared in a memo to department personnel.

The Anthropic Exclusion and Legal Fallout

Notably absent from the update is Anthropic’s Claude. The relationship between the Department of War and Anthropic soured over the company’s refusal to allow its technology to be used for autonomous weaponry or surveillance. While Hegseth previously designated Anthropic a “supply chain risk,” a judge ruled last week that the department acted unlawfully in making that designation.

The initial rollout of GenAI.mil in December was reportedly hasty, with some personnel initially fearing the "I want YOU to use AI" pop-ups were part of a cyberattack until official memos were circulated.