Anthropic, Google, OpenAI, xAI just scored the DoD’s Biggest AI Deal Yet

Anthropic, Google, OpenAI, xAI just scored the DoD’s Biggest AI Deal Yet

On Monday the Pentagon’s AI shop (Chief Digital and AI Office) handed four of the industry’s hottest labs contracts that could total $800 million, aiming to bolt large-language-model smarts onto everything from combat planning to payroll.   

Key Points

  • Four vendors, $200 M ceilings, two-year prototype deals
  • Focus: “agentic” AI workflows for warfighting and back-office tasks
  • Part of CDAO’s commercial-first push as DoD chases billions for AI in FY-26

“The adoption of AI is transforming the Department’s ability to support our warfighters and maintain strategic advantage over our adversaries,” said Chief Digital and AI Officer Doug Matty as the awards dropped, framing the move as a shortcut past slow, bespoke development.  

Each Other Transaction Agreement carries a $200 million ceiling and a two-year window—time for the companies to prototype “agentic” workflows that can read classified data, reason across it, and spit out recommendations inside existing DoD platforms like Advana and Maven Smart System. Earlier this summer OpenAI landed the pilot slot; now Google, Anthropic, and Musk’s xAI join the lineup.  

Google Public Sector says the deal opens its Tensor Processing Units and “Agentspace” orchestration stack to the Pentagon, plus an air-gapped version of Google Distributed Cloud already cleared at IL-6. “These advanced AI solutions will scale agentic AI across enterprise systems,” VP Jim Kelly wrote.  

OpenAI is packaging its most capable models in a secure enclave dubbed “OpenAI for Government,” pitching use cases from proactive cyber defense to trimming health-care paperwork for troops. Anthropic will field its Claude Gov family, built for classified networks, and lean on risk-forecasting research to spot adversarial misuse. xAI, fresh off controversy over Grok’s rawer public chatbot, unveiled “Grok for Government”—a suite it says every federal agency can now buy through the GSA schedule.  

The Pentagon’s spending surge follows Task Force Lima’s 2024 report urging a “commercial-first” sprint and comes as the FY-26 budget asks for multibillion-dollar AI and autonomy funds.

But agentic AI raises new questions about how much autonomy military systems should have. While the Pentagon says these tools will focus on "mission areas" like logistics and data analysis, the line between support functions and combat operations isn't always clear in modern warfare.

For the companies involved, these contracts represent more than just revenue—though $200 million is pocket change for firms like Google and OpenAI, which generates $10 billion annually. The real value lies in establishing relationships with the Pentagon and proving their AI can handle high-stakes, classified work.

The broader implications extend beyond military applications. As these companies build AI systems tough enough for national security work, those capabilities inevitably flow back into civilian products. The internet, GPS, and countless other technologies followed this same military-to-consumer pipeline.

Whether that's reassuring or concerning probably depends on how you feel about AI systems that can act independently in situations where the stakes couldn't be higher.

Chris McKay is the founder and chief editor of Maginative. His thought leadership in AI literacy and strategic AI adoption has been recognized by top academic institutions, media, and global brands.

Let’s stay in touch. Get the latest AI news from Maginative in your inbox.

Subscribe