AI Boom Driving US Inflation Up 50 Basis Points by Year-End

AI hardware constraints fuel price increases

The artificial intelligence boom is generating measurable inflationary pressure in the United States, with effects significantly more pronounced than in other developed economies, according to new analysis from Goldman Sachs.

The investment bank estimates that AI is currently lifting core personal consumption expenditures inflation—the Federal Reserve's preferred gauge—by approximately 20 basis points annually in the US. That impact is projected to more than double by the end of 2025, reaching a 50 basis point contribution to core PCE inflation.

By contrast, Canada, Australia, Europe, the United Kingdom, and Japan are expected to see an average increase of just 10 basis points. Goldman economist Megan Peters characterized AI-driven inflation as predominantly "a US story."

Three waves of AI inflation

Goldman's research identifies three distinct channels through which artificial intelligence is pushing consumer prices higher.

Memory chip prices have surged amid intense demand for AI hardware. An 8GB DDR5 memory module now averages $148, more than triple the $35 price point from the same period last year, according to computer hardware service firm Pangoly. Goldman projects US software and accessories inflation will peak before the end of 2026, with prices growing at a 30% year-over-year rate in November. Memory inflation hits the US harder because software and accessories represent roughly 1% of PCE inflation domestically, compared to less than 0.5% in other developed nations.

Software pricing is climbing as companies bundle AI capabilities into existing products. Microsoft's decision to raise prices for its 365 suite after incorporating its Copilot AI tool exemplifies this trend. Software accounts for a larger share of core inflation in the United States than in any other developed economy Goldman measured.

Electricity costs represent the third pressure point. The average price per kilowatt-hour in US cities reached $0.19 in May, up 27% since May 2022, Bureau of Labor Statistics data shows. Data centers currently consume 6% of total US power demand, a figure Goldman expects to nearly double to 11% by decade's end.

Why it matters

The concentrated inflationary impact in the United States complicates the Federal Reserve's policy calculus as it attempts to maintain price stability while supporting economic growth. The analysis suggests AI's near-term economic effects differ markedly from the long-term productivity gains many forecasters anticipate. While Goldman expects AI to eventually reduce inflation—though potentially less dramatically than past technology cycles like the 1990s internet boom—the timing of that transition remains uncertain. For now, supply constraints in critical components are creating real price pressures that policymakers must navigate.

These findings were first reported by Business Insider, based on research from Goldman Sachs economist Megan Peters.