A recent remark by a prominent Harvard economist regarding the potential use of artificial intelligence in government analysis has triggered a broader debate among economic policy veterans over the risks of overreliance on automated tools in high-stakes government work.
Jason Furman, who previously chaired the White House Council of Economic Advisers, told a media outlet that if he were to return to his former post today, he would rely on AI tools rather than staff economists for much of the council’s day-to-day analytical tasks. The statement drew swift reactions across online policy circles, prompting pushback from other former government economists and administration officials who argued that such a perspective undervalues the critical role of human judgment in policymaking.
Jared Bernstein, who served as CEA chair under President Biden, emphasized that the work of the council is highly bespoke and involves explaining complex data flows to busy leaders who require more than standard market analysis. Writing on Substack, Bernstein noted that leaders look to understand how specific reports fit into a broader political agenda, narrative, and schedule.
Within a White House environment, Bernstein argued that the human skill of reading the room is paramount, expressing skepticism that artificial intelligence can replicate that capability. He pointed out that this conclusion stems in part from listening to the creators of the technology itself.
Claudia Sahm, a former Federal Reserve economist, raised concerns about the pipeline for future leadership, a worry echoed across consulting and investment banking sectors. Sahm pointed out that junior staff members from Furman’s tenure went on to lead major research institutions, run Treasury economic policy, direct Federal Institute initiatives, serve in subsequent administrations, and publish influential research.
Pointing to the career paths of former junior staffers, Sahm argued that models cannot replicate the professional development and institutional growth fostered by early-career positions, asking where future senior leaders will originate if those entry-level roles are eliminated.
Anna Wong, chief U.S. economist at , acknowledged the empowering nature of the technology for senior officials who can now execute ideas independently. However, Wong cautioned against longer-term structural risks, warning that relying on artificial intelligence because of a depleted pipeline of human experts could lead to a scenario where machines generate the foundational ideas.
Describing that trajectory as a potential disaster that social policy must eventually address, Wong highlighted the dangers of bypassing human expertise in senior advisory roles.
Conversely, Casey Mulligan, who served in the first Trump administration’s CEA and later as chief economist at the Department of Health and Human Services, offered a different perspective based on firsthand experience building AI tools in two federal offices.
Mulligan stated on social media that the implications for human labor are not necessarily straightforward. Having used AI tools to conduct economic research faster and more effectively, Mulligan reported that the office actually required more economists rather than fewer due to a surge in demand for analytical outputs, noting a similar pattern in the small business sector.
This phenomenon, where labor-saving innovations ultimately increase the demand for labor, is recognized in economics as Jevons Paradox.
Addressing the challenge from a managerial perspective, Stanford economist Neale Mahoney urged senior professionals to focus on building organizational structures where humans and artificial intelligence complement one another while actively preparing younger workers for a technology-driven landscape.
