GAO publishes AI oversight agenda, with plans to adopt more AI itself
The congressional watchdog will examine AI deployment, national-security risks and policy effectiveness while building its own technical workforce and testing tools.
Loading page…
The congressional watchdog will examine AI deployment, national-security risks and policy effectiveness while building its own technical workforce and testing tools.
Listen to this story
GAO’s 2027–2032 plan makes its own capacity part of the test: the watchdog must develop technical expertise and securely evaluate tools as it scrutinizes federal AI governance. It sets out planned reviews and internal adoption, not findings that specific systems have failed or confirmation that new tools are already in use. Annual performance reporting will give Congress and the public a stated checkpoint for tracking whether those commitments are carried out.
The plan calls for national-security assessments covering adversarial AI use, autonomous systems and the integrity of AI-enabled decisions.
GAO’s Innovation Lab is intended to securely test tools for audit and business processes as the agency expands use of AI, data analytics and cloud technology.
GAO plans to streamline hiring for AI, data science and cybersecurity specialties, while training current staff in analytical, cybersecurity and AI skills.
The Government Accountability Office is putting AI on both sides of its work: a subject for federal oversight and a tool for its own operations. Its fiscal 2027–2032 strategic plan, released October 16, sets priorities for examining AI risks and rules while building the watchdog’s technical workforce and expanding its use of emerging technology.
GAO is Congress’s independent, nonpartisan watchdog, providing analysis and recommendations on federal programs and spending. The new plan places technology within that wider assignment rather than treating AI as a separate mission. It organizes the agency’s work around three goals: independent analysis for Congress, better federal performance and accountability, and stronger internal capabilities. The previous plan used four goals.
The AI oversight agenda covers deployment and the policies governing it. GAO intends to assess how AI is used and overseen, the risks involved, international competitiveness, and whether policy and regulatory frameworks are effective. That makes the rules surrounding adoption part of the review, alongside the technology being deployed.
National security receives its own AI objective. The plan calls for examining federal efforts to address adversarial use, autonomous systems and the integrity of AI-enabled decisions. These are subjects for future assessments, not findings that particular systems have failed. They sit alongside GAO’s broader responsibilities for reviewing military readiness, intelligence management and emerging threats.
Other technology objectives extend the agenda beyond AI models. The plan identifies three additional areas of scrutiny:
GAO also plans technology assessments and other work explaining emerging science and its policy implications to Congress. That gives the agenda a forward-looking role as well as an oversight role: helping lawmakers understand new developments, not only evaluating how agencies manage technologies they have already adopted.
We must be agile as congressional needs and priorities rapidly change and federal agencies reimagine how they carry out their missions using emerging technology.
Orice Williams Brown, acting U.S. comptroller general, in the strategic plan
Inside GAO, the plan points toward broader agency-wide use of AI and data analytics, along with cloud technologies. MeriTalk reported that the agency intends to use its Innovation Lab to securely test tools that improve audit and business processes. The internal technology effort therefore covers both oversight work and the operations supporting it.
Staffing is another part of that effort. GAO intends to streamline hiring for specialties including AI, data science, cybersecurity and emerging technologies. It also plans training to strengthen employees’ analytical, cybersecurity and AI skills. The workforce agenda combines recruiting technical expertise with developing the capabilities of existing staff.
The framework was developed with input from congressional clients. GAO says it sought comments from oversight and other congressional committees, advisory bodies, inspectors general, state and local audit organizations, and independent experts. It also requested comments from the Congressional Budget Office and Congressional Research Service, reviewing and incorporating feedback as appropriate.
GAO says it will track progress annually through its Performance and Accountability Report. The plan connects strategic objectives to performance goals, providing Congress and the public a structure for following implementation. That annual reporting is the stated checkpoint for an agenda whose commitments span external technology reviews, internal adoption and workforce development.
Loading discussion...
Join the conversation
Explain whether firsthand experience strengthens scrutiny or calls for extra safeguards.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.