Fed’s Jefferson Says AI Is Lifting Demand Before Productivity, Flags Job Risk
The Takeaway: The path of inflation, still well above the Fed’s 2% goal, is the more important signal for setting interest rates, Vice Chair Philip Jefferson said, ahead of the central bank’s December meeting. He expects new technology to lift the economy’s capacity substantially but gradually. A day earlier, Governor Michael Barr warned that the gains from artificial intelligence could be shared more narrowly than those of past breakthroughs, with labor’s slice of national income already at a record low.
Artificial intelligence is lifting US spending and some prices before productivity, Federal Reserve Vice Chair Philip Jefferson said, warning that it could also displace workers faster than they adapt.
Speaking at the University of California, San Diego on Friday, Jefferson compared today’s AI boom with the computer age of the late 1980s, when the economist Robert Solow observed that computers showed up everywhere except in the productivity statistics. The Vice Chair asked the same question of the new technology: how much businesses and workers actually use it, and whether that use is showing up in the data.
His answer was that it is not yet. The evidence so far, Jefferson said, is consistent with a contribution to aggregate productivity that is limited, but with the potential to grow quite substantially, in line with past technological breakthroughs. He added that the size and timing of the gains remain highly uncertain.
Adoption is wide but shallow
Surveys find that 40% to 50% of working adults use generative AI in their jobs, Jefferson said, but only a modest fraction use it every day. Census Bureau data cited by Jefferson show roughly 25% of US firms using AI in at least one business function.
Use varies widely by industry. The share is closer to 40% in knowledge-intensive sectors such as professional services and information, and around 10% in mining and in leisure and hospitality. “AI is not truly everywhere,” Jefferson said, describing an early stage in a long cycle of innovation and deployment.
Over 2020 to 2025, labor productivity growth averaged 2.2%, the Vice Chair noted, a step up from 1.5% over the previous business cycle. That surge began before popular AI chatbots debuted, however, and likely reflects several factors, including a pickup in new business formation.
Firm-level studies point in different directions. One survey of senior managers he cited estimated that AI has raised sales per employee by just 0.5% over three years. Jefferson said firms are still experimenting, using AI in only a few functions, and that gains in one area can bring new costs in another: some firms report that AI-written code contains more errors and needs more debugging.
Demand moves first, supply later
The larger effect for now is on demand, the Vice Chair said. Expectations of future productivity gains are already raising spending significantly, and a substantial part of the rise in equity prices over the past few years reflects enthusiasm about the technology. That added wealth appears to be lifting household spending.
Companies are building and equipping data centers on a very large scale, driving rapid growth in business investment. This spending is putting upward pressure on resource use and on the general price level, Jefferson said. Because investment demand is concentrated in some sectors where supply has struggled to keep pace, such as memory and logic chips, prices of specific goods have risen steeply, he said.
He expects that pressure to ease over the medium term, as chip supply catches up and productivity gains from adoption materialize. Memory chip makers have announced a large expansion of production capacity, a footnote to the speech says.
The Vice Chair also turned to the neutral rate, the interest rate that neither stimulates nor restrains the economy. If AI permanently raises productivity growth, firms would want to invest more and households might save less, and the neutral rate would likely rise. In that case, any given level of the federal funds rate would effectively be less restrictive. Jefferson cautioned that the link between productivity and real interest rates has historically been noisy.
The risk to jobs, and Barr’s warning on inequality
Jefferson said the transition could be painful for some workers and businesses. Firms that lag in adoption could fall behind, and workers in tasks that AI performs more efficiently, such as parts of software development, may face job displacement.
One risk to his baseline view, he said, is that “the new technology could displace workers much faster than they can adapt and retrain, which could result in an increase in unemployment.” Monetary policy can affect cyclical conditions such as a temporary downturn, the Vice Chair noted, but not the structural factors that set long-run employment.
Governor Michael Barr took the distribution question further at Yale University on Thursday. He said the share of national income going to labor, as opposed to profits, interest and other returns to capital, fell to a record low of 52.8% in the second quarter of 2026. After World War II, it was fairly steady between 60% and 65%. Some of the decline reflects how income has been measured, Barr said, but even after adjusting for that, the downward trend is clear.
AI could be what some researchers call capital-biased innovation, Barr said, with returns going mainly to owners of capital rather than to workers. Barr, who has examined in earlier speeches this year whether AI could displace more workers than past technologies, turned this time to the risk that its benefits could be more narrowly shared than those of earlier technological revolutions. He warned that the advantage of the leading AI firms may prove unprecedented.
So far, Barr said, there has been relatively little displacement of workers, and competition among AI firms has kept access fairly broad. But there are signs that AI may substitute for labor in some sectors, especially for younger, less experienced workers. He said that “we should be prepared for the possibility of serious short-term disruptions in the labor market,” adding that now is the time for society to consider how to address these challenges.
Inflation is the signal that matters for rates
Turning to the economy, Jefferson said output grew at a 3.4% annual pace in the third quarter and that he expects growth next year roughly in line with potential, the pace the economy can sustain without stoking inflation. The unemployment rate settled at 4.1% in October, Jefferson noted, a level he views as consistent with maximum employment.
Inflation is the concern. The personal consumption expenditures price index, the Fed’s preferred gauge, rose 3.6% in the year to September, Bureau of Economic Analysis data released on 29 October show. Excluding food and energy, the rise was 3.1%, Jefferson noted. Data received so far for October remain consistent with well above-target inflation, the Vice Chair said, and he remains concerned that higher energy prices could lead to a persistent rise in inflation more broadly.
Jefferson voted with the majority in October to keep the federal funds rate unchanged at 3.75% to 4.00%, as we reported on 28 October. Two members, Lisa Cook and Beth Hammack, dissented in favor of a quarter-point increase.
Data since then, he said, show solid activity and persistent price pressures. “Accordingly, I see the trajectory of inflation as providing the more important signal for determining the appropriate stance of monetary policy,” he said, adding that further adjustments should depend on the data, the evolving economy and the balance of risks.
That ranking was absent from his 1 October speech, in which he saw the risks to inflation as tilted to the upside and tied future adjustments to the data, the evolving economic picture and the balance of risks.
He did not say whether the Fed should raise rates at its next meeting. Governor Christopher Waller said on 10 November that he leans toward a quarter-point increase in December, as we reported. He added that a half-point increase should be considered if October consumer price data proved similar to recent monthly readings. Vice Chair for Supervision Michelle Bowman said the same day that there is still work to do to bring inflation closer to 2%.
In The Edge’s reading, Jefferson’s ranking of inflation above a stable labor market as the guide for policy leaves a December increase open without committing to it, while his and Barr’s remarks show Fed policymakers treating AI as a source of price pressure today and of supply gains only later.
What to watch: The Bureau of Economic Analysis publishes October PCE inflation on 25 November. The Federal Open Market Committee meets on 8 and 9 December, with new economic projections, and then on 26 and 27 January 2027, dates the Fed lists as tentative.
Method: quotes are from speeches posted on federalreserve.gov by Vice Chair Philip Jefferson on 13 November 2026, Governor Michael Barr on 12 November 2026, and Governor Christopher Waller and Vice Chair for Supervision Michelle Bowman on 10 November 2026; inflation data are from the Bureau of Economic Analysis release of 29 October 2026.
Sources: Board of Governors of the Federal Reserve System, Bureau of Economic Analysis, The Edge.



