AI trades off with other federal R&D
Christie Lawrence & Sean Cordey 20, concurrent Master of Public Policy and Juris Doctorate 2024 candidate at Harvard Kennedy School and Stanford Law School and recipient of Harvard Kennedy School’s John F. Kennedy Fellowship. She previously worked as a management consultant for Oliver Wyman where she focused on regulatory risk and global data protection for financial institutions. She also worked at the Department of State. Her research with the Belfer Cyber Project includes AI policy, cybersecurity, US-Russian cyber relations, and US foreign policy in the digital age. a dual Master degree in International Affairs candidate at the Fletcher School of Law and Diplomacy and the University of St. Gallen and was a Belfer Cyber Security Project research assistant. He previously worked for the Swiss ministry of foreign affairs in Washington D.C. and currently works as a researcher for the Cyber Defense Project at the Center for Security Studies (CSS) at the ETH Zürich. The Cyber Project Belfer Center for Science and International Affairs Harvard Kennedy School. "The Case for Increased Transatlantic Cooperation on Artificial Intelligence" August. https://www.belfercenter.org/sites/default/files/2020-08/TransatlanticAI.pdf //pipk
AI Funding: Although the Administration has pledged to increase (non-defense and defense) AI-related spending and absolute AI R&D budget numbers have increased, there are concerns that these numbers may not accurately reflect development. First, as AI-related expenditures have increased, the budget for all government R&D has decreased.158 For example, the President’s budget request for cuts in R&D at NSF, NIH, DOE, and other agencies, would force these government entities to prioritize AI R&D to the detriment of other, potentially equally useful R&D.159 Second, without full transparency about the procedures undertaken to re-classify projects as AI-related, it is not possible to fully credit the supposed increase in AI-related R&D to new AI projects. One analysis by Bloomberg Government of the Pentagon’s FY2020 budget found that approximately 27% of the legacy AI-related activities had not included any AI components or descriptors in the previous year budget.160, 161 Observers have suggested that the DoD was partaking in “AI-washing,” or exaggerating the increase in its AI-related R&D to meet government imposed objectives. 162 The US government will need to be careful that initiatives to enhance AI innovation do not foster a zero-sum competition between AI and other S&T research but instead foster genuine innovation.