Pentagon uses AI to write congressionally mandated reports

The Department of Defense is using generative AI to draft congressionally mandated reports and other personnel documents, Pentagon Chief Technology Officer Emil Michael said on June 12 at the Hudson Institute, and the department says GenAI.mil usage rose to 1.5 million personnel in June 2026.

How is the Pentagon using AI to write mandated reports?

GenAI.mil, the Defense Department’s bespoke platform, has been available since December 2025 and initially used Google Cloud’s Gemini for Government; Emil Michael described using it to turn a task that would "otherwise take 200 hours of staffing time" into work done "in five hours." The department made AI tools available across all six military branches, and the Pentagon CTO said the number of Department of Defense personnel using commercial AI tools through GenAI.mil climbed from 80,000 in December 2025 to 1.5 million in June 2026. Jacob Glassman, deputy assistant secretary of defense for science and technology foundations, told an April 23 audience that a short-staffed team he directed to "use GenAI.mil" returned a week later saying it was "the best report we’ve written in the past five years." The department also uses AI for personnel evaluation reports, commendation citations, and counseling statements.

What safeguards or review processes are in place?

The source text does not describe specific review or vetting processes for AI-generated reports, and the government has not disclosed how it verifies accuracy before sending documents to Congress. The article notes the Pentagon has struggled to deliver mandated reports in a timely way: the Government Accountability Office showed the number of required reports rose from just over 500 in 2000 to more than 1,400 by 2020, and the GAO previously found identifying and assigning reports could take between three and six months. The piece also cites broader institutional caution: outside organizations such as KPMG withdrew an AI-produced report after researchers found numerous AI-generated errors and false claims, a cautionary example of how automated drafting can introduce mistakes if not adequately vetted.

Why it matters

The surge in AI use intersects with Congress’s oversight role and a rising workload for the Pentagon. Mandated reports are a core accountability mechanism for how the military spends taxpayer dollars, and the department has requested a $1.5 trillion budget for fiscal year 2027. If AI-generated drafts contain errors or mischaracterizations, those mistakes could undermine oversight. At the same time, leaders inside the department present the tools as a practical response to an exploding reporting burden and staffing constraints.

What to watch

Watch whether the Department of Defense publishes formal review procedures for AI-generated congressional reports and whether it discloses contract details tied to its May 1 agreements to deploy AI tools on classified networks. On May 1 the department announced agreements with eight companies to extend AI for lawful operational use on classified networks, naming SpaceX, OpenAI, Google, Nvidia, Reflection AI, Microsoft, Amazon Web Services, and Oracle; the government has not divulged how much it is paying under those contracts. Tracking any public guidance or audit results that show how AI drafts are checked will reveal whether the department balances speed gains with the accuracy oversight requires.