Discussions are spreading across central banking circles that artificial intelligence (AI) could raise the neutral interest rate itself—the benchmark for monetary policy—if it lifts productivity across the economy. The argument is that if AI raises potential growth, boosting corporate returns on investment and demand for funds, the neutral rate—the theoretical real rate that neither stimulates nor restrains the economy—could come under upward pressure. The Bank of Korea is also in the middle of re-estimating the neutral rate, drawing attention to how AI-driven productivity changes will factor into the future path of its policy rate.
At the recent International Monetary Fund (IMF)-World Bank Annual Meetings in Bangkok, Thailand, the impact of AI on interest rates emerged as a key agenda item. Shin Hyun-song, governor of the Bank of Korea, raised the possibility that AI could push up the neutral rate, saying on October 13 that “whether AI will raise the natural rate of interest is an interesting question.” He added, “In a simple growth model, the result is that the higher the growth rate, the higher the interest rate.” Lesetja Kganyago, governor of the South African Reserve Bank, also voiced the difficulty of pinning down the level, saying, “We know the neutral rate has risen, but we don’t know how high it has gone.”
The neutral rate serves as a compass for monetary policy. If AI raises productivity and lifts potential growth, companies’ expected investment returns rise and demand for funds increases, putting upward pressure on the neutral rate. In that case, the same level of policy rate would have a weaker tightening effect than before—meaning higher rates would be needed to contain inflation.
That is a significant variable for the Bank of Korea’s future rate decisions. The Bank of Korea raised its policy rate twice in a row, in July and August, lifting it from 2.50% to 3.00% a year. The possibility of further hikes remains amid the semiconductor boom and inflationary pressure. Still, some point out that the current semiconductor export boom does not necessarily mean a structural rise in South Korea’s potential growth rate. If AI actually lifts South Korean productivity and raises the neutral rate, the appropriate level of the policy rate and the policy path could also change. The Bank of Korea is currently re-estimating the neutral rate; in research published in 2024, it estimated the long-term real neutral rate at -0.2% to 1.3% as of the first quarter of 2024.
The bond market also sees prospects for further hikes reflecting the possibility of a higher neutral rate. Kim Ji-na, an analyst at Eugene Investment & Securities, cited the rise in South Korea’s GDP gap and the possibility of a higher neutral rate as reasons for another hike, leaving open the possibility that the policy rate could rise to 3.75% next year if inflationary pressure persists.
Still, higher AI productivity does not necessarily mean a higher neutral rate. According to research published by the Bank of Korea in August, if AI raises productivity, capital returns and investment demand increase, putting upward pressure on the neutral rate, but if labor displacement crimps consumption and investment, the neutral rate could instead come under downward pressure. The analysis suggests the direction of the neutral rate depends on the size of the productivity gains and how much of those gains flow through to household income and consumption.
Counterarguments that empirical evidence is still lacking are also substantial. In research published in August, Jens Christensen, a researcher at the Federal Reserve Bank of San Francisco, and Glenn Rudebusch, a senior fellow at the Brookings Institution, said various U.S. neutral-rate estimates have risen by about 1 percentage point since 2020, but they found no statistical evidence that AI-related news explains that increase. That means further verification is needed on whether AI’s productivity improvements will structurally lift the neutral rate.
The market is watching the fact that the Bank of Korea’s judgment is intertwined with short-term rate decisions. The median of the dot plot (rate projections six months out) presented by the Bank of Korea’s Monetary Policy Board in August was 3.25% a year, 0.25 percentage points above the current level. If the re-estimation of the neutral rate comes out higher, the dot plot and the market’s view of the terminal rate could be adjusted accordingly. Conversely, if the conclusion is that the AI effect is limited, the debate over the ceiling for tightening could also shift.
Ultimately, the key question is whether AI investment shows up in actual productivity statistics. If productivity gains are confirmed, a higher neutral rate would translate into weaker tightening intensity, strengthening the argument that higher policy rates are needed to stabilize prices. On the other hand, if AI-driven labor displacement and weaker consumption become pronounced, the neutral rate could fall instead, leaving central banks to manage risks in both directions at once.