(CNN) — AI is raising prices for Americans – and not just electricity bills.
Major technological innovations carry the potential to transform economies by creating opportunities, jobs and even new industries while supercharging productivity and growth.
However, that promise of longer-term gains often is preceded by short- and medium-term pain.
In the case of artificial intelligence, that has included job losses, slower wage growth, widening wealth inequity and, especially in recent months, higher inflation.
Recent data shows that the gargantuan interest and investment in AI adoption (estimated to be around $750 billion for this year alone) have pushed a variety of prices higher, lifting overall inflation in the process.
“The higher inflation means that households must spend just over $375 more to purchase the same goods and services as they did this time last year due to AI’s inflationary impact,” Mark Zandi, chief economist at Moody’s Analytics, wrote in an email to CNN.
The good news: AI is still just a minor contributor (an estimated 0.2 percentage points) of overall inflation, and the impacts are currently limited to a handful of categories.
But the not-so-good news: AI is pushing inflation higher and further compounding longstanding affordability concerns in the process. Plus, these price pressures aren’t expected to go away anytime soon, and they very well could broaden.
That potential dynamic has Federal Reserve officials, including the central bank’s new chairman, on alert.
Here’s a look at where AI has already shown up in inflation and where it could crop up next.
Electricity
AI data centers can have voracious appetites for energy (notably electricity and water), and the rapid expansion of these monoliths threaten to strain grids and drive prices up further.
That’s largely because demand is outrunning supply.
Data center facilities can be built or expanded at double or triple the pace of new electricity generations systems needed to serve them, PJM Interconnection, America’s largest grid operator, noted recently.
Combine those needs with retiring coal plants, increased electrification needs, extreme weather and aging infrastructure, and it further widens the gap between supply and demand.
“Data centers are demanding huge amounts of power, and that’s tending to crowd out the electricity available to distribute to residents; it’s also led to wholesale electricity prices being bid up, because data centers are willing to pay the price that providers ask them for, and that ends up also raising the prices for residential electricity costs,” Pooja Sriram, US economist at Barclays, told CNN.
US Consumer Price Index data shows that residential electricity prices rose about twice as fast in 2025 as compared to the average seen in years prior, she noted.
And through the first five months of this year, electricity prices were climbing even faster than in 2025, Bureau of Labor Statistics data shows.
“I think that is one of the clearest imprints of AI data center demand driving up residential electricity costs,” she said.
Electricity prices unexpectedly fell 1% in June but continue to outpace overall inflation and are up 4% from a year ago, the latest CPI data shows.
Memory chips
The data centers’ appetites, however, aren’t fully sated with power alone. The massive buildouts also led to a surge in demand for memory chips, which has benefited manufacturers handsomely, Sriram said.
“The issue is not just the demand; the issue is the supply side for those memory chips has been very constrained,” she said.
The trillions of dollars chasing AI-related components now have storage and memory suppliers prioritizing their wafer-manufacturing capabilities toward high-bandwidth and high-speed (and highly profitable) memory products commanded by data centers, she said.
“What that does has basically diverted (the production of) the memory chips that you need for consumer products toward very specific high-performance memory chips that the data centers need,” she said.
The pricing pressures of these and other components have been most evident at the producer level. The Producer Price Index chart semiconductor and other electronic component manufacturing industry looks like a hockey stick.
As of June, that category’s wholesale prices were up 26% from a year ago – a stark shift from June 2025 when prices were falling 0.8% on an annual basis, PPI data shows.
“Why this matters for the end consumer is, at the end of the day, our laptops, computers, iPhones, iPads all have some sort of memory chip embedded in that hardware, and those chips have become quite expensive,” she said.
Computer hardware and software
Late last month, Apple hiked the prices for some of its most popular products by roughly 20%. In a statement, the company noted that AI data centers created an “extraordinary surge” in demand for memory and storage.
Sony upped the price of its PlayStation console earlier this year, and Microsoft last month raised the price of its Xbox consoles by about 25% for similar reasons.
“The entire consumer electronics industry is struggling with the current components crisis, but the effects are particularly hard on consoles,” Microsoft wrote in a statement.
Computers and related hardware have typically been a highly deflationary product category in the CPI: Because of technological advancements, consumers can get more bang for their buck. (For example, a $1,500 computer in 2025 was likely more powerful than a $1,500 year-ago model, and as such, the BLS treats this as a price drop)
For the first half of 2026, however, computers and related products have experienced price inflation, BLS data shows.
“We’re in the early innings of these consumer price pressures and the pass-through from higher producer prices, higher import prices and greater demand, especially for AI-led investment,” said Gregory Daco, chief economist at EY-Parthenon.
Adding AI features in business applications also affects the price of software. For example, Microsoft raised personal Office 365 prices by 43% in February (30% for a family plan) after keeping them steady for a decade. The new feature: Copilot, Microsoft’s new AI tool.
Construction costs, wages
Data centers also are impacting the supply of other key construction inputs such as copper and electrical wiring, as well as workers.
“If you’re looking for data that was conclusive (about AI’s effect on the economy and inflation), albeit a bit more subtle, you would look and see whether wages in construction were going up more than wages in the rest of the economy,” said Thierry Wizman, Global FX and rates strategist at Macquarie Group.
“Because if in fact there is upward pressure, straining resources of the economy because of the AI data center buildout, you would see it in wages as well – specifically in the wages of labor that would be working these projects,” he added.
So far, the available national-level wage data is showing a “robust divergence” between the construction sector and the aggregate, he noted.
Regional data could prove even more telling, he said, noting the importance of tracking construction wages in areas with a high concentration of data centers. (That will take more patience, however, as that more localized data is lagged due to collection and modeling needs).
“We’re having a problem with housing in the country these days; people talk about it as being unaffordable,” Wizman said. “It could be the case that the fact wages in construction have been rising a lot is putting upward pressure on houses as well.”
The-CNN-Wire
™ & © 2026 Cable News Network, Inc., a Warner Bros. Discovery Company. All rights reserved.