OpenAI is teaming up with the U.S. Department of Defense on a new pilot program to build “frontier AI” capabilities. The contract could be worth up to $200 million, with $2 million paid immediately, according to the DoD’s daily contract list.
The DoD states OpenAI will develop prototype AI tools to tackle national security issues across both warfighting and enterprise areas. OpenAI announced this under its new “OpenAI for Government” initiative aimed at bringing its advanced AI tech to U.S. agencies.
The project will explore how frontier AI can improve DoD admin tasks—like helping service members apply for healthcare—and boost cyber defense efforts. OpenAI is careful to say its tech won’t be used to develop or control weapons, following its usage policies. 
The U.S. government has been ramping up AI partnerships. In December, the DoD gave Anduril Industries a $100 million contract for military tech. Anthropic is also working with AWS and Palantir to supply AI to defense and intelligence agencies.
OpenAI CEO Sam Altman told former NSA chief Paul Nakasone in April the company is “proud to and really wants to engage in national security areas.”
Last week, OpenAI Chief Product Officer Kevin Weil and ex-Chief Revenue Officer Bob McGrew joined the U.S. Army Reserve as lieutenant colonels. They’ll work with a new detachment advising the government on AI in the military, alongside execs from Palantir and Meta.
OpenAI is also pushing Project Stargate, which plans to invest $500 billion in U.S. AI infrastructure.
This marks a deepening of ties between OpenAI and the U.S. government—blurring lines between commercial AI and national security.
Defense Department contract details: DoD contract list
OpenAI announcement: OpenAI for Government
