Citi Spots AI Inflation Twist That Could Hand New Fed Chair Warsh Room to Ease

Markets have spent months debating whether artificial intelligence will one day crush inflation. Right now, the data tell a different story. Surging demand for memory chips and servers has pushed specific tech prices higher. Those increases show up in core inflation readings. And analysts at Investing.com report that Citi sees this as a potential opening for Federal Reserve Chair Kevin Warsh to take a more dovish stance.

Shortages bite hard. The “Computer Software and Accessories” sub-index within core CPI has climbed 13.9% year-over-year. For 25 years it mostly fell. No longer. AI-driven data center buildouts have created a supply crunch in memory. Prices reflect that scarcity. Yet the Fed’s preferred gauge, core PCE, tells a slightly different tale. April readings showed core CPI at 2.8% year-over-year against core PCE at 3.2%. The gap matters.

Citi’s Analysis Highlights Narrow Price Pressures

Citi analysts argue the memory price spike represents relative price changes in a narrowly defined good. Business demand, not broad consumer spending, drives it. “Rather than a generalized consumer inflation, the increase in memory prices is an increase in the relative price of a narrowly defined good that is in short-supply due to business demand,” they wrote. This distinction appeared in the Fed’s April meeting minutes. Some officials observed that recent price increases in the information technology sector contributed to higher inflation. A few added that software category gains “may not be good predictors of future overall inflation.”

The equity boom adds another layer. Higher stock valuations tied to AI lift portfolio management fees. Those fees flow directly into core PCE calculations. So AI doesn’t just raise some input costs. It inflates asset values that feed back into official price measures. But again, this channel remains narrow. Not the broad wage-price spiral that keeps central bankers awake.

Warsh took the oath as Fed chair days ago. President Trump selected him. The new leader has long viewed technology as a disinflationary force over time. He argued in a Wall Street Journal op-ed that AI boosts productivity and competitiveness. Lower rates could accommodate that shift. Now Citi hands him data that supports treating current inflation distortions with caution. If the pressures prove temporary and sector-specific, aggressive rate hikes might prove unnecessary. Or even damaging.

But not everyone buys the dovish case. Goldman Sachs economists estimate AI-related price pressures have already added 0.3 percentage points to annual core PCE inflation and 0.1 points to core CPI over the past year. They forecast similar effects ahead. Fortune detailed the bank’s analysis last week. Memory chip prices have soared. Average selling prices for computers and non-Apple smartphones could rise 10% this year. J.P. Morgan Asset Management noted the same dynamic in its research. AI acts as a massive demand shock for resources and capital in the near term even if it promises lower costs later.

And optimism itself creates inflation. The St. Louis Fed ran its dynamic stochastic general equilibrium model on a “TFP news shock” — exactly the sort of AI productivity expectations now circulating. Households anticipate future pay raises and spend more today to smooth consumption. Firms invest early to prepare. “Together, these forces produce an inflationary surge in aggregate demand,” the researchers wrote. Output and inflation rise even though potential output has not yet changed. Monetary policy decides the outcome. Rate hikes contain the surge. Accommodation makes initial jumps larger and risks persistently high inflation if gains disappoint.

Recent Fed commentary reflects the tension. Governor Christopher Waller signaled the central bank would remove its implicit easing bias from the June policy statement. Chair Warsh inherits a committee already shifting parameters. Minutes and speeches show officials divided on AI’s labor market and price effects. Some fear near-term demand pressures outweigh any productivity payoff still years away. Others, including Warsh previously, see AI as justification for lower rates to capture future gains.

Markets price in the uncertainty. The Philadelphia Semiconductor Index posted an 18-day winning streak earlier this year. AI-related equities have propelled major indexes to records despite geopolitical risks and bond sell-offs. Yet producer prices jumped. Oil prices signal further inflation risks. X posts from traders capture the skepticism. One called Citi’s dovish spin “peak banker brain.” Another noted oil prices hint at massive inflation spikes even as Warsh touts AI disinflation.

The long-term case remains intact. Over years, AI should lower production costs, raise productivity, and exert downward pressure on prices. J.P. Morgan Asset Management research agrees AI will likely prove significantly disinflationary eventually. But timing matters. Easing now based on distant productivity would loosen financial conditions immediately while inflation lingers. That mismatch explains why many analysts urge patience.

Warsh faces his first FOMC meeting soon. Data will test his views. If memory prices stabilize and broader inflation metrics improve, the narrow AI distortions Citi identified could indeed provide breathing room. Policymakers might look past the software index spike. They could treat equity-driven fee increases as transient. Yet if capex keeps climbing — Goldman and others project AI investment nearing $1.2 trillion next year — demand pressures could broaden. Electricity costs, construction, and skilled labor shortages already strain sectors beyond chips.

So the Fed must separate signal from noise. Is this inflation the good kind? The kind that reflects investment in future capacity rather than excess demand? Citi thinks so. The minutes hint some officials agree. But models like the St. Louis Fed’s warn that mistaking optimism for realized gains can lock in higher inflation. Warsh has signaled openness to growth-oriented policy. His challenge lies in convincing markets and colleagues that today’s price bumps do not preclude tomorrow’s productivity boom.

Investors watch closely. Rate cut expectations have swung with each inflation print. Bond yields reflect competing forces — AI optimism on one side, energy and capex costs on the other. Equity valuations, especially in technology, embed enormous productivity assumptions. Any sign that Warsh treats AI-related inflation as less concerning could ease financial conditions further. That would reinforce the very equity gains feeding back into PCE.

The setup is unusual. A new Fed chair. Narrow but visible inflation distortions tied to the economy’s hottest theme. Competing research from banks and the Fed system itself. And real-world shortages that refuse to resolve quickly. How Warsh threads this needle will shape policy for the next several years. For now, Citi has given him intellectual cover to look past some of the headline numbers. Whether he uses it remains the next market-moving question.