American military personnel will soon have access to military-grade versions of ChatGPT and Grok as the Pentagon aims to enhance its use of artificial intelligence in everyday operations.
On Monday, the Department of Defense announced that these AI tools will be integrated into its dedicated platform, GenAI.mil, which was introduced in December and initially provided only a tailored military version of Google’s Gemini. Currently, GenAI.mil boasts 1.7 million users out of the Department of Defense's workforce of three million, encompassing both military and civilian employees.
This latest development follows the Pentagon's awarding of significant defense contracts to OpenAI and SpaceXAI, which operates Grok, with each contract potentially worth up to $200 million. Additionally, Google and Anthropic also received contracts during this period.
The military-specific version of ChatGPT, referred to as ChatGPT Mil, is part of OpenAI's initiative aimed at deploying its AI solutions to government employees. Joe Larson, OpenAI's Vice President for Government, stated, “ChatGPT on GenAI.mil will streamline unclassified work within the Department, allowing personnel to allocate more time and resources towards core missions.”
According to a Pentagon press release, “ChatGPT Mil provides users with a familiar commercial interface within a secure environment, specifically designed to meet the needs of military personnel.” This version of ChatGPT, which ignited a competitive landscape when it debuted in 2022, ensures that data remains within its environment and is not used for further model training. Service members will be equipped to utilize both this specialized ChatGPT and Grok for Controlled Unclassified Information (CUI) at Impact Level 5 (IL5), which pertains to sensitive information not classified.
Much of the military's activities involve unclassified tasks with heavy documentation related to human resources, logistics, and contract work—areas where ChatGPT's capability to manage vast amounts of data could be beneficial, according to sources familiar with OpenAI's applications.
As for the military version of Grok, named Grok for Government, the Pentagon elaborated that it will empower personnel with advanced inference capabilities and varying degrees of reasoning (auto, fast, and expert). This tool aims to enable faster and more precise mission execution in various operational settings, ranging from market research for acquisitions to supply chain management for logistics.
The introduction of ChatGPT and Grok aligns with Secretary of War Pete Hegseth's advocacy for integrating AI into the Pentagon's daily processes, reflecting the White House’s AI Action Plan, which emphasizes the necessity for the U.S. to adopt AI aggressively to maintain military superiority.
Shifting towards an "AI-first" approach, the Department of Defense is not only changing its operational methods but also its strategic outlook, according to an official. However, reports from DefenseScoop indicated that the rollout of GenAI.mil was somewhat rushed, leaving many in the workforce unaware of the new platform's launch.
On the launch date of December 9, employees encountered a pop-up on their screens featuring a stylized GenAI.mil logo and the phrase “I want YOU to use AI,” along with a link to a government site. The abrupt nature of this pop-up raised concerns among some officials, who initially suspected it might be a hacking attempt.
Those concerns were alleviated when personnel later saw a memo from Hegseth and videos on social media in which he stated, “The future of American warfare is here, and it’s spelled AI.” Hegseth urged all Department members to log in, familiarize themselves with the platform, and incorporate AI into their workflows as a daily practice.
Notably, Anthropic's Claude was not included in the recent addition of Grok and ChatGPT to GenAI.mil. The AI firm, led by Dario Amodei, publicly fell out with the Department of Defense after Hegseth criticized AI models that "won’t allow you to fight wars," a comment later understood to target Anthropic's request that its technology not be used for autonomous weapons or surveillance during contractual discussions.
In February, former President Donald Trump ordered federal agencies to cease using Anthropic’s technology, a process expected to conclude by the end of this month. Hegseth described the company as a “supply-chain risk,” leading to a lawsuit from Anthropic a month later.
A judge ruled last week that the Department of Defense acted unlawfully in designating Anthropic as a supply-chain risk.
