The same artificial-intelligence boom rattling the stock market this week is hitting Americans in a quieter place: their electric bills. The point was underscored Tuesday, when the U.S. Energy Department announced $17.5 billion in loans to build new nuclear reactors to meet the skyrocketing power demand from massive data centers. Behind that lies a problem households already feel. According to the U.S. Energy Information Administration, residential electricity prices have risen more than 36% since 2020, to 17.44 cents per kilowatt-hour, and are expected to reach 19.01 cents by September 2027 — faster than inflation.

The culprit, in part, is the explosion of data centers — the warehouse-sized buildings of computer servers that power AI. The International Energy Agency estimates data centers accounted for roughly 50% of all growth in U.S. electricity demand last year. The Energy Department says data centers used 4% to 5% of the nation’s electricity in 2024, a share that could nearly triple by 2028. Building the plants and lines to serve them costs money — and much of it lands on ordinary ratepayers.

Here is how, in plain terms. When a giant new electricity user plugs in, the local utility often must build new infrastructure. Under the rules in most regions, those costs are spread across everyone on the system, not just the company that created the demand — so even households that never touch an AI chatbot help pay for it. In the mid-Atlantic grid known as PJM, which covers 13 states, prices have risen dramatically as data-center demand has increased.

The dollars are real. The consultancy PowerLines found utilities requested more than $30 billion in rate increases last year, affecting 81 million Americans, and that power bills have risen about 40% since 2021. A Bloomberg analysis found electricity costs in areas near data centers jumped as much as 267% over five years. One Manassas, Virginia, homeowner told Consumer Reports his monthly bill spiked to $281 in January from about $100 the month before.

There is a striking imbalance in who pays. A Yale Climate Connections analysis found that between 2020 and 2024, residential electricity prices rose about 25%, while commercial prices rose far less and industrial users actually paid lower prices. Families running air conditioners and refrigerators have absorbed steeper increases than the big users driving much of the new demand. In industry parlance, ordinary consumers are “captive ratepayers” because, in many states, they cannot shop for a cheaper provider.

That has made electricity a political flashpoint before November’s midterms. President Trump has embraced AI as a growth engine but increasingly sees electricity prices as a threat, and secured a promise from Microsoft that its data centers would not drive up prices. Major operators including Amazon, Google, Meta and Microsoft have signed pledges to build or buy their own power so the cost does not fall on neighbors.

It would be wrong to pin the entire increase on AI. Analysts note bills were climbing well before the boom, driven by an aging grid, higher gas and equipment costs, coal and gas plant closures, and outdated utility profit models. Goldman Sachs analyst Manuel Abecasis estimated higher electricity prices will add about 0.1% to core inflation through 2027 and warned the drag falls hardest on lower-income households, for whom power is a bigger share of spending.

For investors, the same surge has a flip side: utilities, long treated as sleepy stocks, are being valued for growth as they spend billions to serve data centers and recover the cost from customers. That is the uncomfortable knot at the center of the AI build-out. The technology promises enormous gains, but a large share of its immediate cost is showing up on the monthly bills of households that had no say — a tension now driving policy fights in more than 30 statehouses and shaping the midterm campaign.

JBizNews Desk | New York

© JBizNews.com All Rights Reserved. Reproduction or distribution without written permission is prohibited.