As the federal government increasingly relies on artificial intelligence (AI) solutions to execute a variety of missions, it is critical that auditors rapidly enhance its evaluative capabilities in considering a range of policy, transparency and accountability issues as well as computational challenges given the speed of continuing technological advancements. GAO is undertaking an effort to develop an AI oversight framework. In alignment with principles promulgated under the Generally Accepted Government Auditing Standards (GAGAS), our work will encompass considerations such as assessment criteria, requirements for technical and governance artifacts, and evaluation practices. The overall objective is to assure complete, accurate, and evidence-based evaluation of AI systems.
Current ecosystem of AI governance
Current state of AI oversight
GAO's iterative approach towards developing an AI oversight framework
Chief Data Scientist and Director of Innovation Lab,
U.S. Government Accountability Office