While adoption of artificial intelligence in federal agencies already is producing some positive results and has the potential for much more, better coordination is needed to promote its use and mitigate against risks, according to a Senate report.
The report was filed in support of S-1363, which would put in law the AI Center of Excellence within the GSA while defining certain tasks for it, and would require OPM to identify skills needed in AI-related occupations, assess the level of such skills currently in the federal workforce and prepare three-year and 10-year hiring and workforce plans.
Currently, more than 100 agencies use some form of AI to improve services and the government’s “demand for AI solutions will continue to grow,” said the report. “Estimates show automating routine tasks alone could allow federal agencies to deliver better, faster services; utilize up to 1.2 billion federal employee labor hours for more critical tasks; and, potentially save the federal government up to $41.1 billion annually.”
It adds, “Notwithstanding AI’s potential benefits for Federal agencies, some have raised concerns about the technology’s susceptibility to developing or perpetuating bias, lack of transparency and ability to explain decisions, and the adequacy of AI talent to meet current and future demand.”
It cited a September 2019 White House summit on artificial intelligence in government recommending that agencies incorporate best practices being developed elsewhere and create a way for agencies to share expertise among each other.
The bill, which recently passed the committee level, is now ready for a floor vote.