US Military Deploys Custom ChatGPT, Grok on Secure AI Platform

United States — Ekhbary News Agency

The U.S. Department of Defense (DoD) has rolled out custom versions of Grok and ChatGPT, integrating them into its secure artificial intelligence platform, GenAI.mil. This platform, cleared for sensitive but unclassified information, first launched last year featuring Google Gemini. The Pentagon now reports a significant adoption rate, with over 1.7 million unique users from its more than three million personnel actively engaging with GenAI.mil.

Advanced AI Tools for Military Operations

Starshield AI’s Grok for Government, developed by SpaceX’s secure satellite network, is set to provide immediate productivity gains for military personnel. It promises stronger knowledge continuity and more secure, efficient collaboration. Key capabilities include deep-thinking inference, adaptable reasoning modes, customizable workspaces, and persistent projects with reusable ‘playbooks’ to scale institutional knowledge. The DoD affirms Grok will enable missions to be executed faster and with greater precision, spanning uses from market research for acquisition officials to complex supply-chain management. ChatGPT Mil, conversely, targets document-heavy unclassified work across the Department, encompassing planning, policy, logistics, and administration. This model aims to accelerate routine tasks, thereby freeing up valuable time for critical endeavors. All things considered, this strategic deployment underscores the Pentagon's aggressive push to integrate advanced AI, despite ongoing legal challenges with key industry players.

Vendor Diversification Amidst Legal Disputes

By offering multiple AI models, the department seeks to avoid vendor lock-in, a crucial consideration given its recent disputes with Anthropic. Notably, Claude remains the only major AI model absent from the GenAI.mil lineup. Earlier this year, Anthropic, previously a key partner for classified systems, reportedly encountered an impasse in negotiations with the DoD. Officials had insisted on language allowing Anthropic’s technology for "any lawful purpose," raising concerns about potential uses in domestic surveillance and autonomous weapons. After these discussions faltered, the Trump administration controversially designated Anthropic as a supply chain risk. Soon after, tech giants including OpenAI and SpaceX secured agreements with the Pentagon to deploy their advanced AI capabilities on classified military networks for "lawful operational use." Last week, U.S. District Judge Rita Lin ruled in one of Anthropic’s lawsuits that the government’s actions constituted "unlawful retaliation," violating the company’s First and Fifth Amendment rights. A second related case is still pending in federal appeals court in Washington, D.C.