Story
Nov 5, 2026
OpenAI’s Pentagon Push Gains a Former Army Secretary—and Fresh Scrutiny
OpenAI and Driscoll cast the appointment as a way to help the United States and its allies deploy AI responsibly, while critics see a test of whether rapidly tightening ties between AI firms and the Pentagon can be governed without compromising procurement integrity.
Dan Driscoll’s move from the Pentagon to OpenAI comes after a turbulent exit from government—and at a moment when the AI company is steadily expanding its defense footprint.
Driscoll left his post as U.S. Army secretary in September, after 18 months marked by reported disputes with Defense Secretary Pete Hegseth over promotions and the pace of military modernization. Before departing, he had pushed cheaper battlefield drones, less red tape for drone contractors and a shift toward commercial software rather than bespoke military systems.1
On Monday, OpenAI named him senior counselor for national security modernization. The bespoke role will have him advise the company’s public- and private-sector partners on policy and strategy, with an emphasis on responsible AI use by the United States and its allies. Driscoll framed the task as practical rather than abstract: AI can help institutions “cut through paperwork, make sense of more information, and get the right answer to the right person faster.”1
For OpenAI, the appointment extends a visible march toward the national-security establishment. It added former NSA director Gen. Paul Nakasone to its board in June 2024, later partnered with defense firm Anduril, and secured a $200 million Pentagon contract in June 2025. Its subsequent agreement to place models on secure Pentagon networks permitted “all lawful use,” while retaining stated limits on intentional domestic surveillance.1
Supporters argue that government experience is essential if democratic countries are to compete. Sam Altman said the United States has an “extraordinary opportunity” to help other countries adopt “democratic AI.”1
But Gwanhoo Lee of American University sees a more uncomfortable pattern: hiring prominent former officials can strengthen a company’s government relationships and its business with agencies. He warned that procurement shaped by relationships rather than technological merit “puts us at risk,” and called for clearer oversight and conflict-of-interest rules.1
OpenAI said the parties understand and will comply with post-employment restrictions. The larger question is whether those safeguards can keep pace with AI’s accelerating arrival inside the defense state.1