GAO: Pentagon’s Responsible AI Effort Hobbled by Inactive Council and Staffing Shortfalls

The Pentagon’s effort to put “Responsible AI” guardrails around its growing use of artificial intelligence is at risk because a key coordination council went inactive for months and the office leading the work has been operating with a fraction of its authorized staff, according to a new Government Accountability Office report.

The report, published Oct. 14 and titled “Artificial Intelligence: Actions and Opportunities to Implement Responsible Use in DOD,” was requested by Congress through Senate Report 118-188, which accompanied the fiscal 2025 National Defense Authorization Act bill. The findings come as the Department of Defense is already using AI at scale: GAO said DOD reported 378 AI use cases as of February 2026.

GAO reviewed how the department is implementing and overseeing its Responsible AI guidance, which builds on DOD’s AI Ethical Principles adopted in 2020. The department later defined a Responsible AI approach around six tenets intended to help ensure AI systems are safe and trustworthy. GAO found that AI is already in use across DOD and that most of the 10 defense components it surveyed had taken at least initial steps to set up oversight and apply those tenets in selected use cases.

Still, GAO said the department’s main working-level coordination body faltered at a critical point. The Responsible AI Working Council, run through the Office of the Chief Digital and Artificial Intelligence Officer, stopped meeting for much of the period GAO examined. “CDAO officials acknowledged that the council did not meet from late 2024 until August 2025,” the report said.

GAO said the council resumed meetings in August 2025, but warned it remains vulnerable to another lapse because it still lacks a formal charter laying out its roles, responsibilities and meeting schedule. That gap mattered across the department. Four DOD components told GAO they faced Responsible AI implementation challenges during the time the council was not meeting and that collaboration would have helped. Six of the 10 surveyed components told auditors that sharing Responsible AI resources and best practices would support their implementation efforts.

The other major constraint was staffing. GAO said a 2023 resourcing memo authorized 10 positions for the CDAO’s Responsible AI Division, but as of February 2026 only three of those jobs were filled. According to GAO, CDAO officials told auditors the office was unlikely to fill the vacancies quickly because it was restructuring after an August 2025 realignment under the Defense Department’s Under Secretary of Defense for Research and Engineering.

The staffing shortfall stands out against the scope of the work still underway. GAO said that, according to CDAO documentation, 29 of 64 lines of effort in DOD’s Responsible AI Pathway were still incomplete as of April 2026. In GAO’s view, the shortage of personnel directly affects how quickly the Pentagon can move from policy to implementation. “By addressing current workforce needs for the Responsible AI Division, CDAO could make more progress toward its goal of accelerating AI across DOD,” the GAO product summary said.

GAO made two recommendations to the Defense Department: ensure that CDAO develops and issues a charter for the Responsible AI Working Council, and ensure that CDAO addresses the workforce needs of its Responsible AI Division. The watchdog also included a matter for congressional consideration, saying lawmakers should consider directing DOD to implement earlier GAO recommendations on AI collaboration and workforce issues that remain unresolved.

According to the GAO product page, DOD concurred with both recommendations in comments provided by email. The report leaves little doubt about the central issue facing the Pentagon’s AI push: DOD is moving ahead with hundreds of AI applications, but GAO says the governance and staffing meant to keep that expansion consistent and responsible have not kept pace.