Budget · AI adoption
GAO finds staffing gaps in Pentagon's responsible AI division
The division was tasked with accelerating AI adoption by building trust through assessment, test and evaluation, and monitoring, officials told GAO.
The Department of War's Responsible AI Division had three of its 10 authorized positions filled, officials told the Government Accountability Office.2 The count was as of February 2026; GAO published its report Oct. 14.1–2
Pentagon responsible AI division had three of 10 positions filled
Responsible AI Division staffing, February 2026
Officials reported this staffing count to GAO as of February 2026.
Source: GAO-27-107837, published Oct. 14, 2026. Staffing as of February 2026.
Officials from the Chief Digital and Artificial Intelligence Office, or CDAO, described the division's task as accelerating AI adoption by building trust through assessment, test and evaluation, and monitoring. The department's January 2026 AI strategy re-emphasized accelerating integration and adoption.1
CDAO officials said the office was unlikely to fill vacancies or add staff, citing ongoing restructuring after its August 2025 realignment and the need to account for the new strategy. As of April 2026, CDAO documentation showed 29 of the 64 lines of effort in the department's 2022 responsible AI implementation pathway were incomplete.1
Officials said the Responsible AI Working Council did not meet from late 2024 until August 2025. Officials from four components reported implementation challenges during that period that would have benefited from collaboration. Meetings resumed, but GAO warned that the absence of a charter defining responsibilities and meeting frequency left the council at risk of lapsing again.2
GAO recommended issuing a council charter and addressing the division's workforce needs; the department concurred with both recommendations, which GAO lists as open.1–2 GAO also recommended that Congress consider directing the department to implement its prior recommendations on AI collaboration and workforce issues.2
In its use-case survey response, the Air Force reported that industry partners' proprietary AI models lacked transparency in data, algorithms or model characteristics, hindering objective assessment of responsible AI implementation. GAO cautions that these survey and interview examples do not represent trends across the department.1
Sources (2)
- GAO-27-107837, Artificial Intelligence: Actions and Opportunities to Implement Responsible Use in DOD (report PDF) www.gao.gov
- GAO product page for GAO-27-107837 www.gao.gov