Surging electricity demand and prices have recently been driven by the rapid construction of data centers needed to train and operate artificial intelligence systems. And those power-hungry AI data hubs may create unanticipated consequences for consumers — including increased housing costs.

The 30-year fixed-rate mortgage — the most widely originated home loan in America — was designed with that amortization schedule to provide investors with a hedge against volatility and inflation, and borrowers with an affordable, predictable payment schedule.

The process of qualifying borrowers for that mortgage is straightforward. Underwriters assess borrowers’ creditworthiness and issue commensurate mortgage interest rates, judging affordability — a borrower’s ability to repay their loan — based on income, assets and outstanding debts.

Over the past five years, however, the flexible costs of monthly mortgage payments like property insurance and taxes have risen precipitously, challenging ingrained assumptions about homeownership costs, long-term affordability and the viability of purportedly predictable payments for 30-year mortgage products.

Property insurance costs rose 11.3% year over year in September, according to a report from ICE Mortgage Technology, a division of Intercontinental Exchange Inc., which owns the New York Stock Exchange.

As a portion of borrowers’ monthly mortgage payments, property insurance costs have averaged 70% growth nationally over the past five years, while principal has risen by 23%, interest by 27% and property taxes by 27%.

Insurance and taxes are calculated as a point-in-time metric, treated for the purpose of underwriting as fixed, when in reality those costs are flexible. As a result, escrow accounts through which homeowners often pay those portions of their mortgage have ballooned in recent years, with annual insurance payments rising to 9.6% of average monthly mortgage payments in July.

Spiking electricity costs

Sneaking up on homeowners now are spiking energy bills, not included in monthly mortgage payments or underwriting criteria, but nevertheless an essential monthly housing expense. Analysts warn that electricity costs will climb steeply over the next decade as demand outstrips current and projected new supply, squeezing household budgets.

ICE Climate, another division of Intercontinental Exchange, provided Scotsman Guide with an advance copy of an upcoming white paper highlighting energy cost implications on home affordability, using energy bill estimates as of January 2025 for more than 100 million single-family households across the U.S.

ICE researchers estimate that households living in single-family detached homes spend an average of about $3,300 on energy per year. Even as property insurance costs continue to rise, ICE says energy costs represent the largest “additional” or flexible cost of monthly homeownership expenses — and will likely only keep rising.

ICF International Inc., a U.S.-based technology and consulting firm, projects electricity demand in the U.S. to expand 25% by 2030 from 2023 levels. The consulting firm McKinsey & Company forecasts U.S. power demand increasing up to 3.5% annually through 2040.

As of May, a U.S. household needed to earn $114,000 annually — 70% higher than the $67,000 threshold six years ago — to afford a median-priced home of $431,250, assuming a 30-year fixed-rate mortgage, a 20% downpayment and no more than 30% of gross monthly income spent on housing, according to Realtor.com.

“Variation in energy bills through the year in many parts of the United States suggests that a more complete framework for assessing home affordability — one that includes not only total energy costs, but also variations in these costs through time — should be considered by investors and lenders alike,” ICE researchers note.

That monthly variation in total energy costs can have “profound implications for affordability” due to a feature they share with other rising flexible costs of homeownership: unpredictability. Nearly half of U.S. consumers have less than $500 in savings and more than 78% live paycheck to paycheck, according to ICE.

“A few months of unexpectedly high energy bills could have long-lasting effects on individual households,” the researchers conclude. “A household with an unexpectedly high energy bill one month may experience prolonged financial stress throughout the year.”

AI-fueled energy crunch

Short-term projections from the U.S. Energy Information Administration (EIA), housed within the U.S. Department of Energy (DOE), show U.S. power consumption will hit record highs in 2025 and 2026. Price hikes driven by the demand surge, already underway, vary widely by U.S. region and are projected to outpace inflation through 2026.

The EIA reports average U.S. residential electricity prices rose from 13.16 cents per kilowatt hour in 2020 to 16.48 cents per kilowatt hour in 2024, a rise of 25%. Average prices are projected to swell to 17.91 cents per kilowatt hour through 2026, a 36% jump from 2020.

The DOE reported in December 2024 that data centers consume 4.4% of total U.S. electricity in 2023. The data center share is projected to rise to as much as 12% by 2028.

With homes in roughly 80% of U.S. counties already unaffordable for typical residents by standard housing cost-to-income ratios, according to real estate analytics firm Attom, loving or fearing AI will have no bearing on whether homeowners can avoid the increased expenses that the world-changing technology will inject into housing costs.

In areas of the U.S. proximate to AI data centers that have already been built, Bloomberg reports wholesale electricity costs are as much as 267% higher than five years ago.

“These wholesale commodity costs are passed on to households and businesses on their utility bills, which then include other charges to maintain and expand the network,” Bloomberg reports. “That can affect even customers who aren’t in close proximity to a data center, since their energy relies on the same grid.”

A March research report from the ratings agency S&P Global warned of surging electricity demand: “The next five years pose a major risk of supply and demand imbalance, as data center buildout is expected to go through major development, while near-term supply response is constrained.”

Homeowners fund AI race

The Trump administration released an AI action plan in July outlining government ambitions to stretch the capacity of U.S. power grids while cutting decades-old clean water and air regulations “in a race to achieve global dominance” of AI.

The second of three pillars in the government’s plan, “Building American AI Infrastructure,” begins with a call to “stabilize the grid of today as much as possible,” an initial phase that “acknowledges the need to safeguard existing assets and ensures an uninterrupted and affordable supply of power.” Affordable electricity supply already looks to be falling behind.

Congressional lawmakers have begun to address the impact on rural households, where many data centers have been built. But data centers are being built in all manner of communities across the U.S., especially concentrated along the Eastern Seaboard, meaning the impact of surging electricity costs, fueled by AI’s energy demands, will be felt by homeowners in every voting district.

In a September press release introducing the Unleashing Low-Cost Rural AI Act, U.S. representatives Jim Costa, D-Calif., and Blake Moore, R-Utah, called on the U.S. departments of Energy, Interior and Agriculture to study the effects of AI data center expansions in rural communities.

Costa and Moore cited independent monitoring research showing PJM Interconnection — the largest power grid operator in the U.S., serving roughly 65 million customers in 13 states from Illinois to Washington, D.C. — will charge ratepayers $9.3 billion more “in just one year than they would have without data centers’ electricity demand.”

Traditionally, the Energy Department considers household energy costs affordable if they total less than 6% of a household’s annual income. This annual perspective fails to capture affordability challenges stemming from monthly variations in household energy bills.

A survey from the EIA published in 2018 found that about 1 in 5 U.S. households (approximately 25 million, including renters and owners of multifamily housing units) cut back on essential items like medicine and food to pay energy bills. Another 11% reported keeping their indoor home temperatures at levels that were unhealthy or unsafe — with 7 million households reporting making those tradeoffs every month.

In an era of increasingly unpredictable extreme weather and shifting climate patterns, the impact of rising energy costs on affordable homeownership — and homeowners’ monthly housing outlays — stands to intensify, with homes and grids facing heightened climate risks and the costs of hardening structures and grids rising the longer they are delayed.