The Pentagon has officially launched custom versions of OpenAI’s ChatGPT and xAI’s Grok, according to Defense Department announcements. This rollout extends generative artificial intelligence tools tailored to “warfighter needs” across a workforce of 3 million civilian and military personnel. The new additions integrate directly into GenAI.mil, a centralized, secure portal initially launched in December.
GenAI.mil Platform Adoption and Core Capabilities
The internal generative AI platform has onboarded more than 1.7 million unique users out of the department’s total 3 million personnel, according to the Defense Department. GenAI.mil originally launched with Google’s Gemini. The platform aims to give Department of Defense employees access to commercial frontier models without routing sensitive government data through standard consumer channels. Official military versions of these tools are exempt from the data collection practices common in commercial consumer tech products.
ChatGPT Mil for Unclassified Routine Tasks
ChatGPT Mil, developed through the OpenAI for Government program, delivers a user experience similar to commercial ChatGPT versions. According to the Defense Department, the platform focuses on chat, files, projects, and custom GPTs. The tool supports document-heavy, routine unclassified work such as administrative tasks, logistics, planning, and policy. Additional features will roll out over time.
Grok for Government and Starshield AI Integration
The Pentagon framed the capabilities and potential applications of SpaceX’s Starshield AI version of Grok in broader, more militaristic terms. According to a Defense Department press release, Starshield AI’s Grok for Government provides the warfighter with immediate productivity gains, stronger knowledge continuity, and secure, efficient collaboration. The system utilizes SpaceX’s secure satellite network built on existing Starlink technology. The Defense Department stated that Grok enables the military 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. Grok for Government also includes multiple reasoning modes, customizable workspaces, and reusable playbooks.
Commercial Partnerships and Anthropic Supply-Chain Dispute
This expansion highlights the notable absence of Anthropic’s Claude model. The Trump administration previously designated Anthropic a national security supply-chain risk after the lab refused to grant the Pentagon unrestricted use of its AI tools, insisting instead on safety guardrails. A federal judge struck down that designation, ruling the government unlawfully retaliated, and Anthropic is currently fighting that designation in court. OpenAI took a different path, reaching an agreement that permits its models to be used for lawful purposes while maintaining contractual and technical safeguards. Alongside OpenAI and xAI, the Pentagon has struck deals with Amazon Web Services, Microsoft, Nvidia, and Reflection AI to boost its AI capabilities.
Security Concerns and Unauthorized Shadow AI
The push to expand authorized AI tools comes as the Pentagon wrestles with workforce security risks tied to commercial technology. In June, the Defense Counterintelligence and Security Agency warned that unauthorized “shadow AI” tools could cause data leaks and other security vulnerabilities. Congress has separately ordered the Department of Defense to assess cybersecurity risks stemming from both authorized and unauthorized software.
Frequently Asked Questions
What tools are currently available on GenAI.mil?
GenAI.mil features Google Gemini, OpenAI’s ChatGPT Mil, and Starshield AI’s Grok for Government, according to Defense Department announcements.
Who has access to the Pentagon’s GenAI.mil portal?
The portal is available to the Department of Defense’s 3 million civilian and military personnel, with over 1.7 million unique users already onboarded, according to the department.
What kind of tasks does ChatGPT Mil handle?
According to the Defense Department, ChatGPT Mil handles document-heavy, routine unclassified work including administrative tasks, logistics, planning, and policy.
What security standards apply to GenAI.mil tools?
Tools on GenAI.mil are certified to handle controlled unclassified information at Impact Level 5, according to a Pentagon official.
Keep reading