The Trump administration has identified nearly a billion dollars of research and development spending for non-defense agencies as part of a supplement to the president’s FY 2020 budget request, nearly double what the government spent on this emerging technology three years ago.
U.S. Chief Technology Officer Michael Kratsios said the budget document released Tuesday provided the government’s first agency-by-agency look at non-defense AI spending and serves as part of an executive order President Donald Trump signed in February to prioritize AI in the federal R&D portfolio.
The budget document, Kratsios said, serves as an “important mechanism” for getting a big-picture look at total government spending on AI research.
“This new supplement report demonstrates just how diverse and extensive our efforts are,” Kratsios said at a Center for Data Innovation event. “By breaking down exactly how we are spending our non-defense AI R&D dollars, we can better identify opportunities for future investment, conduct long-term strategic planning across the government and find new opportunities for collaboration between the federal government industry and academia.”
In civilian government, the National Science Foundation, the Energy Department and the National Institutes of Health stand out as some of the top funders of AI research.
The updated budget request shows that overall government spending on AI research is on track to nearly double, compared to just a few years ago.
In 2016, all agencies, including the Defense Department, spent about $1 billion on AI research.
The defense community, meanwhile, has accelerated its own investment in AI research through established institutions like the Defense Advanced Research Projects Agency and newer projects like DoD’s Joint AI Center, which the Pentagon stood up last year.
Military spending on AI alone now approaches $1 billion, Krastios said, but DoD keeps information on its overall AI spending classified.
Anthony Robbins, Nvidia’s North America public sector vice president, said the federal government’s increased spending on AI research put to rest industry concerns that the president’s executive order would amount to an unfunded mandate for agencies. However, he said government research funding would need to accelerate for the U.S. stay on the cutting edge.
“A billion [dollars] is certainly a great thing … but it’s not nearly enough,” Robbins said, pointing out, for context, that agencies spent more than $47 billion on fraud, waste and improper payments in fiscal 2018.
Jackie Medecki, the director and managing attorney of Intel’s U.S. AI and health care policy, agreed that government, together with its partners in academia and industry, should consider a “moonshot” initiative for “major spending” on AI research.
“Honestly, the private sector is very alluring because there’s just more money there, but there are folks who want to do funding and want to be in academia, so if we had more funding and more grants and more research money there, there will be more happening there, too,” Medecki said.
The Trump administration has already looked at ways for government, private industry and academia to brainstorm on AI research through public-private partnerships.
At a White House AI summit Monday, Anil Cheriyan, director of Technology Transformation Services at the General Services Administration, discussed standing up an AI Center of Excellence within GSA, aimed at delivering “transformative improvements in government operations.”
Tuesday’s budget document outlines several strategic administration goals, including improving the “safety and security of AI systems so that they operate in a controlled, well-defined, and well-understood manner,” and developing shared public datasets to further test AI tools.
In order to reach those goals, the National Institute of Standards and Technology released a roadmap last month outlining the government’s role in setting up a regulatory framework for AI research.
NIST’s report marked the federal government’s first major effort to provide clarity and guidance to agencies looking to adopt AI technology. The report also highlighted a need to balance regulation with giving agencies and their private-sector partners the flexibility to continue researching new breakthroughs in AI technology.