The U.S. Department of Defense has integrated ChatGPT and Grok from OpenAI into its internal generative AI platform, making it available to approximately 3 million civilian and military personnel. The Pentagon is accelerating the adoption of cutting-edge commercial models into government systems, while emphasizing the isolation of sensitive data and ensuring the security of internal usage.
GenAI.mil Extended Model Integration
This platform is named GenAI.mil, and it was already integrated with Google Gemini in the early stages. The Ministry of National Defense states that the goal of the new version is to allow employees to use mainstream large models without going through ordinary consumer-grade product channels, thereby reducing the risk of sensitive government information leaking out.
According to official statements, the military version does not adopt the common data collection methods used in consumer-grade products. This means that the relevant tools are more suitable for handling routine tasks that are not classified but involve a large volume of files and have fixed processes.
ChatGPT Mil First, cover the documents and planning.
The Ministry of National Defense stated that ChatGPT Mil is provided by the OpenAI for Government projects. The user experience is similar to that of the commercial version ChatGPT. The initial features include dialogue, file processing, project management, and customization GPT.
Currently, this tool is mainly used for administrative, logistical, planning, and policy-related tasks in unclassified information scenarios. The official also stated that additional features will be added in the future.
Grok focuses more on task execution.
In contrast, the Department of Defense's description of Grok places more emphasis on the efficiency of military tasks. An official press release states that Grok for government use will enhance the immediate productivity of frontline personnel, their ability to sustain knowledge, as well as their capability for safer and more efficient collaboration.
The Department of Defense also mentioned that this set of tools can be used in a variety of task environments, including market research analysis for procurement personnel and supply chain management for logistics personnel. Unlike ChatGPT Mil, which focuses on document processing, the description of Grok is more aligned with task execution and cross-scenario collaboration.
Anthropic does not appear on the list.
There is no Claude with Anthropic in this newly added list. Reports mention that the Trump administration previously listed Anthropic as a supply chain risk entity, and the related designation is still under legal dispute in court.
The background of the dispute is that Anthropic once refused to provide unrestricted model usage rights to the Pentagon and insisted on retaining certain security measures. Subsequently, the Pentagon turned to collaborating with other technology companies to continue advancing the development of AI capabilities.
Additional information:In addition to OpenAI and xAI, the US Department of Defense has also recently reached partnerships with Amazon Web Services, Microsoft, NVIDIA, and Reflection AI in order to expand its AI infrastructure and tool capabilities.
