Pentagon Uses AI to Draft Congressional Reports, Officials Say

The US Department of Defense has begun using generative AI to write reports mandated by Congress, according to statements from senior Pentagon officials at recent public events.

Pentagon Chief Technology Officer Emil Michael told attendees at a Hudson Institute event on June 12 that AI tools can draft congressional reports in five hours that would otherwise require 200 hours of staff time. "I have to report to Congress every year on this thing," Michael said. "Let me load all the papers onto it and have it draft me a congressional report."

The Pentagon has made AI tools available to all six military branches through its GenAI.mil platform since December 2025, starting with Google Cloud's Gemini for Government. Usage has grown from 80,000 personnel in December to 1.5 million in June 2026, Michael said, out of a total Defense Department workforce of approximately 3.5 million.

AI-generated reports described as 'best in five years'

Jacob Glassman, deputy assistant secretary of defense for science and technology foundations, provided additional detail during an April 23 event. According to coverage by DefenseScoop, Glassman told a short-staffed team responsible for a congressionally mandated report to "use GenAI.mil, do the best you can." The team reportedly returned a week later claiming the AI-generated report was "the best report we've written in the past five years." Glassman did not identify which report.

The Department of Defense faces a significant reporting burden. The number of congressionally mandated reports increased from just over 500 in 2000 to more than 1,400 by 2020, according to the US Government Accountability Office. The process of identifying reporting requirements and assigning them to appropriate teams can take three to six months, according to a 2023 GAO report.

Why it matters

Congressional reports serve as a primary accountability mechanism for how the military uses taxpayer funds. The Pentagon has requested an unprecedented $1.5 trillion budget for fiscal year 2027. If AI-generated reports contain errors or mischaracterizations—a problem that has plagued other organizations using similar tools—it could undermine congressional oversight at a time of historic defense spending. Consulting firm KPMG recently withdrew an AI report after researchers identified numerous AI-generated errors and false claims.

The Pentagon has not disclosed what review processes exist to verify the accuracy of AI-generated congressional reports. Military personnel are also using generative AI to write performance evaluations, commendation citations, and counseling statements, according to Small Wars Journal.

In May, the Defense Department announced agreements with eight AI companies—including SpaceX, OpenAI, Google, Nvidia, Microsoft, Amazon Web Services, and Oracle—to deploy AI tools on classified networks. The contracts notably exclude Anthropic, which was reportedly blacklisted after refusing to allow unrestricted use of its Claude models for autonomous warfare and mass surveillance.

These details were first reported by Ars Technica.