Watchdog finds federal government trailing in crucial AI needs

Key Points

  • Federal agencies are failing to meet deadlines and build their management of artificial intelligence (AI).
  • Government Accountability Office (GAO) has made multiple recommendations to 19 agencies to better implement federally-required AI strategies.

Many federal agencies are falling behind on deadlines and efforts to improve their management of artificial intelligence (AI), even as the technology’s usage expands within the federal government. This was revealed in a report launched by the Government Accountability Office (GAO) this week.

The GAO report states, “key efforts to strengthen management of AI have missed deadlines and are not yet completed,“. The GAO made dozens of recommendations to 19 federal agencies to enhance AI protocol implementation. These recommendations include developing strategies about AI technology use, maintaining inventory to consist of required information, and issuing guidance on AI application.

The GAO discovered that out of 23 agencies, 20 have reported around 1,200 current or planned application areas that could be benefited or resolved by AI. Activities such as border camera data analysis and drone image analysis were included in the examples provided. Agencies like NASA and the Departments of Commerce, Energy, Health and Human Services, and State had the highest reported AI usage in the fiscal year 2022.

However, only five agencies could provide comprehensive information about their AI use cases. The remaining 15 had either incomplete or inaccurate data; some were even reported to lack mandatory data points like the life cycle of the AI and the releasability of AI instances. The report highlights, “Without accurate inventories, the government’s management of its use of AI will be hindered by incomplete and inaccurate data“.

AI continues to be utilized by the federal government, and its application is expected to surge significantly in the future. Simultaneously, the federal government and Congress are considering how to address and exploit the potential advantages while also recognizing the risks of the technology. There has been numerous legislative initiatives to regulate AI and several bipartisan meetings discussing the same with tech leaders.

There is a continued push for more standards and regulations in the AI landscape, with a bill that enforces standards for AI programs in the agricultural sector recently introduced. Moreover, the U.S. has collaborated with 45 other countries to execute a declaration on the “responsible military use” of AI.