Are Consumers Paying Twice for AI?
Data Centers, Electricity Demand, and What Happens When Grid Costs Meet Community Voice
March 2, 2026. — Susan Greene
As artificial intelligence grows into everyday life — powering search, creative tools, business software, and automation — the physical infrastructure behind it is growing too. Data centers, the facilities that run these services, are among the fastest-expanding electricity users in the United States.
At the same time, electricity prices have been rising for households in many regions — driven by fuel price volatility, transmission upgrades, wildfire mitigation, and infrastructure modernization.
Layered onto these longstanding cost pressures is a new driver of demand: data centers supporting AI and cloud services.
AI Data Centers and Electricity Consumption
AI systems require enormous computing power — and continuous electricity. According to the Pew Research Center, U.S. data center electricity consumption has surged amid the AI boom, with growth expected to continue as generative AI expands
E&E News reports that data centers’ share of total U.S. electricity consumption could double by 2030, largely driven by AI workloads and high-performance computing.
“Data centers’ share of U.S. electricity consumption could double by 2030.”
Christa Marshall — E&E News
Unlike many industrial facilities, AI data centers operate 24/7 and require intensive cooling systems. This constant load increases pressure on generation capacity and accelerates the need for transmission upgrades — costs utilities typically recover through regulated rates.
Capacity Markets, Forecasts, and Consumer Costs
Building more data centers doesn’t just mean putting up new buildings and power lines. It also changes how the electric grid decides how much electricity will be needed in the future — and how much that electricity will cost.
PJM Interconnection is the organization that manages the power grid across 13 states and Washington, D.C., the busiest and largest wholesale energy market in the United States. It tries to make sure there will be enough power plants available to meet demand, especially during the hottest and coldest days of the year. To do that, it looks years ahead and estimates how much electricity people and businesses will use.
When large data centers are added to those forecasts, PJM assumes much more power will be needed. That signals that more power plants must be built or kept running. A review by SemiAnalysis found that when projected data center growth was included in these estimates, the price utilities must pay to secure future electricity supply went up significantly.
Those higher costs don’t just affect tech companies. They are built into utility bills — meaning households and small businesses help pay for the extra power needed to run data centers.
Infrastructure Costs and Ratepayers
Meeting higher demand often requires new substations, reinforced transmission lines, and additional generation resources. The Union of Concerned Scientists (UCS) found that roughly $4.3 billion in transmission connection costs associated with data center expansion in the PJM region were passed through to local customers, with over 95% of identified projects shifting connection costs onto ratepayers.
“Over 95% of the projects identified passed all of their transmission connection costs onto local people’s electricity bills, totaling $4.3 billion in costs.”
Daela Taeoalii-Tipton — Union of Concerned Scientists
UCS has also reported that data center growth is already contributing to higher electricity bills in some areas.
These are not abstract phenomena but structural outcomes of how regulated rate structures recover capital costs.
Negotiated Deals and Transparency Challenges
Large tech companies often negotiate customized power contracts with utilities. Reporting from NBC News shows that many such agreements include nondisclosure provisions, limiting public understand.
Citing CNN, analysts and lawmakers have focused on whether these arrangements might inadvertently shift costs to residential customers.
Transparency matters because electricity infrastructure and rate structures affect everyone, regardless of whether they use a particular cloud service.
Communities Push Back — and What Happens Next
Across the country, communities are beginning to assert their voices in data center siting and energy planning.
In Pennsylvania, grassroots resistance has emerged to dozens of proposed data center projects. Residents and local officials have raised concerns about electricity demand, industrial land conversion, heavy water use, and diesel backup generators tied to planned facilities.
Some municipalities have rejected rezoning proposals after sustained public input.
The challenge is that when local voices influence planning outcomes, some corporate actors may seek alternative routes.
In Iowa, decisions about whether a massive industrial facility can be built are not made in Washington or at the state capitol — they are made locally. Counties control zoning, land use rules, and the conditions developers must meet before construction can begin.
In Linn County, after residents raised concerns about water use, noise, light pollution, and infrastructure strain, local officials drafted a detailed ordinance specifically addressing data center development. The proposed rules would require companies to complete an in-depth water study and negotiate a binding water-use agreement before breaking ground. They would also impose strict limits on noise and lighting, require a 1,000-foot buffer between data centers and residentially zoned property, hold developers financially responsible for damage to roads during construction, and mandate contributions to a community betterment fund.
“We are trying to put together the most protective, transparent ordinance possible,” Kirsten Running-Marquardt, chair of the Linn County Board of Supervisors, told nearly 100 residents at the first public reading of the draft in early February.
“We are trying to put together the most protective, transparent ordinance possible,”
Kirsten Running-Marquardt, Chair of the Linn County Board of Supervisors
But after encountering resistance under this local framework, Google sought annexation — shifting the proposed project into a different jurisdiction. County leaders characterized the move as an attempt to bypass the locally crafted rules and the public process that produced them.
This sequence highlights tensions that can emerge when community engagement is robust but grid and economic pressures push large investment decisions forward.
National Energy Policy and Cost Debates
The question of how to balance data center growth, grid capacity, and consumer costs has also surfaced in national discourse.
InsideClimateNews recently reported on a federal proposal to address data center power costs as part of broader energy affordability commitments. Though framed as a national pledge, the discussion reflects a growing recognition across policy arenas that the intersection of large corporate energy users and consumer bills is a live issue.
Such proposals underscore that energy cost concerns now extend beyond regional markets into broader public debate — even as the mechanics of price formation remain rooted in market rules and regulatory design.
A Fossil-Heavy Grid Meets Rising Demand
The broader U.S. grid remains heavily weighted toward fossil fuel generation. According to Spectrum Local News, in 2025 roughly three-quarters of U.S. electricity production came from natural gas, coal, and nuclear sources, while renewables like wind and solar accounted for about 18% despite rapid growth.
As AI data centers expand demand, questions about how quickly renewable sources can scale to meet new loads — and how grid planners integrate distributed energy resources — remain central to debates over affordability, resiliency, and environmental impact.
Are Consumers Paying Twice?
Many households now pay subscription fees for cloud-based and AI tools. At the same time, they also pay electricity bills that may reflect infrastructure, capacity pricing, and transmission costs arising from serving large corporate loads.
If infrastructure investments needed to serve AI data centers are recovered broadly through rate structures — while large customers negotiate bespoke utility deals — households can carry costs in multiple ways.
This is not about ideology; it is about how cost allocation, market design, and transparency affect affordability.
The Bottom Line
AI and data centers are integral to the modern economy. They power essential services in healthcare, education, finance, and more. But electricity is a shared public system whose planning, pricing, and infrastructure decisions ripple far beyond corporate campuses.
To make sure consumer voices count — and that households are not left bearing the hidden costs of rapid demand growth — policymakers, regulators, and communities must focus on:
- Transparent utility agreements
- Clear cost allocation rules
- Inclusive local planning processes
- Capacity market design that reflects public interest
- Integration of renewable energy to meet new demand sustainably
Consumers may choose whether to subscribe to AI tools. They do not choose how grid costs are structured — which is why visibility into those structures matters.
Because in the context of electricity, we all share the grid — whether we subscribe to the cloud or not.