A report released by the Government Accountability Office (GAO) has revealed that federal agencies in the United States have identified over 1,200 potential uses for AI, with more than 200 already employed in various capacities. The report emphasises the federal government’s need for responsible AI management to mitigate risks, achieve intended outcomes, and avoid unintended consequences.

The GAO report highlights that approximately 69% of the AI use cases identified by federal agencies are focused on science and internal management. NASA leads the way regarding planned and actual uses of AI, with 390 planned uses, followed closely by the Department of Commerce with 285 uses. However, three agencies, namely the Department of Housing and Urban Development, the Nuclear Regulatory Commission, and the Small Business Administration, did not report any potential uses for AI.

According to the report, there are currently 282 AI applications already used by federal agencies, among which 123 have been utilised for over a year. The report highlights NASA’s use of AI for a global surveillance program studying volcanoes and the Commerce Department’s application of AI for satellite-based fire detection and prediction.

Aside from analysing the potential and current uses, the GAO report also examined the extent to which federal agencies have implemented AI-related policies and guidance. The AI in Government Act of 2020 and the executive orders issued by former President Trump and President Biden would be the benchmarks in this process of implementation of AI, according to The Hill. The report states that ten out of the 23 agencies reviewed by the GAO have fully implemented all AI-related requirements. Additionally, twelve agencies are at different stages of implementation, while the Nuclear Regulatory Commission was exempted. Reportedly, the Office of Management and Budget (OMB) and the Office of Science and Technology Policy (OSTP) have yet to meet their AI-related requirements fully.