The United States is likely to bear the brunt of a global wave of inflation driven by AI, according to a new study from Goldman Sachs.
The Wall Street giant crunched the numbers on how the boom in artificial intelligence could cause consumer prices to rise around the world, thanks to supply constraints that are pushing up prices for key components such as memory chips and semiconductors. Megan Peters, an economist at the bank, said the United States was likely to suffer the worst inflation impact.
Goldman estimates that AI boosts core consumer spending inflation (the Fed’s preferred measure of inflation) in the U.S. by about 20 basis points a year. Inflationary pressures are expected to more than double by the end of the year, pushing core PCE upward by 50 basis points.
This is more than the expected rise in core inflation in other developed countries, with Canada, Australia, Europe, the UK and Japan expected to see an average rise of 10 basis points.
“While not completely negligible, these effects are well below the 50 basis points peak we estimate for U.S. PCE, suggesting that most of the AI-driven inflation is a U.S. story,” Peters said.
Peters categorized the impact of artificial intelligence inflation into three different “waves” that could impact the economy.
memory price
Bloomberg/Getty Images
The price of memory chips is rising as demand for AI hardware increases. For example, the average price of an 8GB DDR5 memory module rose to around $148 last week, more than triple the average price of $35 for the same week last year, according to data from computer hardware services company Pangoly.
Goldman said it expects U.S. software and accessories inflation to peak by the end of 2026, with prices rising at a 30% year-over-year pace in November.
Memory inflation is likely hitting the United States harder than other countries, the bank added. Approximately 1% of PCE inflation in the United States is attributable to software and accessories, compared to less than 0.5% in other developed countries.
software price
Thomas Trutschel/Photothek via Getty Images
As more companies bundle software with AI tools, the price of software is also rising. The bank cited several examples where companies have increased prices, such as Microsoft increasing the price of its flagship 365 bundle after incorporating its AI Copilot tool.
Software accounts for a larger share of core inflation in the U.S. than in other developed countries measured, Goldman said.
electricity bill
fhm/Getty Images
Energy is also a major bottleneck for AI trading, given the amount of electricity expected to be needed to power data centers.
The average price for a kilowatt-hour of electricity in U.S. cities rose to $0.19 in May of this year, an increase of about 27% since May 2022, according to the Bureau of Labor Statistics.
Goldman estimates that data centers will account for about 11% of total U.S. electricity demand by the end of 2010, up from 6% now.
Energy prices are also rising due to supply concerns caused by the Iran war. West Texas Intermediate crude has erased most of the gains of the past few months, but is still up 25% year-to-date.
Forecasters say the productivity benefits of AI will eventually lower inflation, but it’s unclear how long the immediate price spike will last before the technology’s disinflationary effects take hold.
Goldman said in an earlier note that it expects AI to be disinflationary over the long term, but that the technology may not be as disinflationary as past technology cycles, such as the internet boom of the 1990s.