New York City bans Generative AI for younger students. What is GenAI?
Mayor Zohran Mamdani announced this week that his administration would ban generative artificial intelligence usage for a majority of students in New York City schools.
Americans have a love-hate relationship with artificial intelligence.
About half of U.S. adults report using AI chatbots, according to Pew Research. A separate Reuters/Ipsos poll found 53% American adults fear the technology could cost them or someone in their household a job.
AI may be best known for automating repetitive work, but it’s also changed the way people approach therapy, the job search and how they manage their money. It’s helping doctors identify breast cancer risks and banks detect financial fraud.
At the same time, AI faces fierce opposition from workers who fear it could take their jobs, artists who worry it could erase their livelihoods and activists concerned about the environmental consequences stemming from the large-scale buildout of data centers, the physical infrastructure needed to support the technology. It’s become a hot-button issue in this year’s midterm elections.
Under President Donald Trump, the United States has gone all in anyway.
“Whoever has the largest AI ecosystem will set the global standards and reap broad economic and security benefits,” the White House said in the introduction to its AI Action Plan, adding “our Nation will win, ushering in a new Golden Age of innovation, human flourishing, and technological achievement for the American people.”
Goldman Sachs Research estimated in August that U.S. investment in the technology will total just under $600 billion in 2026. A BestBrokers report found nearly two thirds of the S&P 500 is now tied to the AI economy. And a JP Morgan analysis found that AI-related capital expenditures added 1.1% to gross domestic product growth in the first half of 2025 – outpacing American consumer spending’s contribution.
As the nation leads in AI investment and model development, Americans are questioning who will benefit most from the AI buildout and how it will affect their wallets.
Has AI driven up inflation?
While many factors are behind the U.S. economy’s stubborn inflation, the AI buildout may be contributing.
Increased demand for things like semiconductors, memory chips, and computer hardware have led to higher prices on those items.
A Federal Reserve Bank of Minneapolis analysis of data from the Personal Consumption Expenditures Price Index, the Fed’s preferred measure of inflation, found that AI-driven demand has pushed up prices in one category called “video and information processing equipment.” Through July 2026, prices were up 12.2% year over year in that category. In comparison, prices in that category fell 6.5% per year from 2015 to 2019.
Overall, core PCE, the Bureau of Economic Analysis’ way to measure price changes while excluding volatile food and energy costs, rose 3.3% over the year in July. The analysis found that the shift in AI-driven demand is responsible for 0.4% of it.
“As the AI investment boom drives demand for memory and other computer hardware, the spillovers to goods prices appear to be at least as large as tariffs in keeping core inflation high,” the analysis said.
In simple terms, inflation remains above the Fed’s 2% annual target, and rising costs related to the AI buildout are one reason why.
Is AI driving up consumer prices?
The PCE index is just one way to measure inflation.
The Labor Department’s Consumer Price Index showed that while prices ticked up 0.1% in July from the prior month, annual inflation cooled to 3.4%.
The report, which also breaks down cost changes by category, features a section called “computers, peripherals and smart home assistant devices,” the best proxy for AI-related prices, according to Dow Jones. From June to July, those AI-related prices rose 3.5%. That increase was faster than the rate for all items combined for the entire year ending in July.
“A lot of computer software and accessories have seen price increases, and because of the price increases, a lot of these businesses are rolling out cost price increases to the consumer,” Shikha Jain, a Simon-Kucher partner and lead of the consumer sector for North America, said. “Memory chip and RAM prices have increased five times because of this AI infrastructure demand, and I think that also has an impact on anything that has any kind of memory component.”
For American consumers, that means they may see higher price tags on things like computers, smartphones, gaming consoles, and tablets. Former Apple CEO Tim Cook told The Wall Street Journal in June that the surging cost of memory and storage chips have made price increases “unavoidable.”
While the AI buildout mostly affects consumer electronics for now, its impact on prices could expand, according to Skanda Amarnath, Employ America executive director and former research analyst for the Federal Reserve Bank of New York.
For example, he said although he hasn’t seen proof of it yet, cars could get more expensive, as they also require memory chips.
“If one supply chain has a higher demand for a particular input, and that outstrips demand from another source, then prices go up,” Amarnath said.
Are data centers to blame for rising electricity bills?
Data centers were responsible for about 4.4% of the nation’s electricity consumption in 2023, according to the Department of Energy’s Lawrence Berkeley National Laboratory. It estimates data centers’ share of U.S. electricity consumption could rise to between 6.7% and 12% by 2028.
While the buildout of data centers increases energy demand, which can push up prices, electricity costs have also soared in recent years due to extreme temperatures and as utility companies replace aging infrastructure.
Data centers’ impact on your electric bill might depend on where you live. A 2025 Carnegie Mellon University and North Carolina State University study found that data centers and cryptocurrency mining could lead to an 8% increase in Americans’ average electricity bill by 2030. For consumers living in major data center hubs, like central and northern Virginia, it estimates their bills could rise by 25% in that time.
Hard-pressed to win favorability with voters over the cost of living, Trump announced a “Ratepayer Protection Pledge” earlier this year. Companies who take the pledge agree to “protect American consumers from price hikes” due to the AI buildout and to “lower electricity costs” for households in the long run. It asks companies to build, bring or buy new power supplies, invest in local job creation and, “whenever possible,” make their energy resources available to communities to prevent blackouts and shortages.
“I’m very proud of it. It was my idea. They’re gonna build their own electric producing facility and that electricity that they have left over is going to the grid,” Trump told reporters in August.
While it’s non-binding, Google, Microsoft, Meta, Oracle, xAI, OpenAI and Amazon have taken the pledge. Governors from 23 states also signed on in July.
“It’s not as if the signatories to these things are actually being held liable for anything,” Amarnath said, adding it is unclear what authority Trump has to enforce such a pledge. “It’s a good aspiration, but I think if you really were serious about this, you probably would have to develop some frameworks for saying this is the accounting for this stuff, this is how we decide that this data center is imposing this cost and needs to pay for it.”
Could AI lead to lower prices over time?
Before taking over as Federal Reserve chair earlier this year, Kevin Warsh said AI will be “a significant disinflationary force” in a November op-ed for The Wall Street Journal.
ConnectOne Bank Founder and CEO Frank Sorrentino explained Warsh’s likely logic in May: AI is expected to make work faster and cheaper, potentially leading to an increased supply of goods and services. Through the laws of supply and demand, prices could fall as a result.
Amarnath gave another example, one in which a consumer may choose to use AI to help them file their taxes rather than pay for a tax preparation service.
“There are some tax prep providers who charge a decent price for them,” he said, adding under this scenario, “they’ll either have to cut their prices aggressively, or they have to find some new ways to provide premium value. Premium value has to be understood in terms of actual unlocks for your tax refund.”
Could AI-driven productivity lead to higher wages?
In the op-ed, Warsh also said increases in productivity should drive “significant increases” in workers’ real take-home wages.
“A 1-percentage-point increase in annual productivity growth would double standards of living within a single generation,” he wrote.
Amarnath isn’t so sure.
“Is it the case that more productive societies can pay their people better? Yes,” he said. “Is it the case that we see when productivity rises, real wages also rise? No. …Even if you assume that productivity is going to increase, it’s not a given that’s going to show up as better conditions for workers.”
In recent months, Warsh has not made such grand statements publicly. Rather than making predictions, he asked questions about AI’s impact when speaking at a closely watched economic symposium in Jackson Hole, Wyoming, on Aug. 28.
“Will the application of AI cause a significant, sustained rise in productivity across the economy? And if so, when?” Warsh asked. “Will token usage be complimentary or competitive to labor?”
He said those were questions for one of the five task forces he created at the Fed focused on productivity and jobs, which will offer policymakers recommendations for changes at the central bank later this year.
Reach Rachel Barber at rbarber@usatoday.com, follow her on X @rachelbarber_, and subscribe to her newsletter “Making More of Your Money” here