Here's how much the AI boom is contributing to inflation, according to Goldman Sachs calculations

By Jules Rimmer

Goldman believes core inflation will revert toward the Fed's 2% target in 2027

The most recent PCE data for August was released September 30 and showed a 3% increase year over year, compared to the Fed's 2% target.

Some officials at the Federal Reserve have blamed surges in demand created by the artificial-intelligence boom for amplifying inflationary pressures. But Goldman Sachs economists argue this overstates the case and that the effect is peaking.

The core personal consumption expenditures index is the Fed's preferred measure of inflation, and the most recent data showed it measured 3% year-over-year for August. Goldman estimates the contribution of the AI phenomenon to be about 45 basis points, or 0.45%.

By 2027, Goldman anticipates core PCE falling closely towards the Fed's 2% target, chiefly because tariffs and the Iran energy shock are fading in potency.

A Goldman Sachs report, published Wednesday and authored by Jessica Riedels and David Mericle, examined recent comments by Fed governor Christopher Waller, for example, and sought to establish just how much impact AI spending, in its various forms, is having on underlying inflation.

The note scrutinized four distinct channels by which AI might contribute to pricing pressures: software and accessories, consumer electronics, electricity prices and other cost spillovers like industrial metals and construction wage growth.

AI has lifted software and accessories inflation by boosting demand for memory and also by higher prices for software products that now incorporate AI features. Riedels and Mericle find numerous ways in which these extra costs have been miscaptured or exaggerated and conclude that "most of the 20 basis points contribution to core PCE from this category reflects mismeasurement."

The increased cost of electronic inputs in computers and consumer electronics goods is estimated to make up around 15 basis points in total to core PCE inflation until late in 2027.

GS expects AI Cost pressures on consumer electronics prices to contribute roughly 15bp to core PCE inflation over the next year

Surging demand for power from data centers certainly has impacted inflation, raising energy costs for companies and household electricity bills and adding 5 to 10 basis points to both core and headline inflation statistics.

Fourth, the price of industrial metals used in the construction of data centers have leapt materially and the Goldman economists calculate this lifted the core inflation number by 5 basis points. However, Goldman finds little evidence of increased wages for construction workers in states where data centers are built.

GS expect the impact of AI price pressures on core PCE inflation to rise from 45bp today to 50bp at end-2026 before falling to 10bp by end-2027 as memory and computer price growth slows

So the 45 basis points of indirect AI pricing pressure that Goldman finds now is expected to peak around 50 basis points at the end of 2026 and then to decline gradually to just 10 basis points by the end of 2027. By next year, the effect is forecast to moderate, Riedels and Mericle suggest, because memory and computer prices should be growing at a significantly slower pace.

-Jules Rimmer

This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.

(END) Dow Jones Newswires

10-14-26 0718ET