Who Should Pay for the Electricity That Powers AI?
May 6, 2026 by
As artificial intelligence (AI) and large language models like ChatGPT become more ubiquitous in American life, “tech giants such as Google, Meta, Microsoft, and Amazon are predicted to spend $364 billion this year to accelerate the construction of new data centers” across the United States.1 Data centers are facilities that house servers, storage systems, networking equipment, and associated components that are needed to supply the massive amount of computing power required to train and run these AI models.2
Background on Data Centers
A side effect of the expansion of data centers is the massive amount of electricity they require, which is equivalent to hundreds of thousands of residential homes or “the electricity demand of small cities.”3 A December 2024 report by the Department of Energy found that in 2023, data centers used 4.4% of all U.S. electricity; by 2028, their consumption was predicted to increase to approximately 6.7% to 12% of total U.S. electricity.4 In addition to their consumption of a large share of the grid, data centers “require expensive upgrades to the electric grid” such as new, stronger transmission lines, substations, and power lines.5
The current business model of the utility industry, which has been consistent for over 100 years, is “spreading costs through everyone.”6 Anytime utilities had to improve or expand their systems, or anytime there was a severe weather event and they needed to fix infrastructure such as powerlines, they spread the cost of those repairs not to the families most directly affected but rather to everybody. They deem such practices as the fairest approach they could take.
However, in the new age of AI, this policy means that “households and smaller companies could be stuck footing these mounting bills” from the data centers of billion-dollar tech companies. A Bloomberg analysis of wholesale electricity prices for tens of thousands of locations across the country found such a reality: “Electricity now costs as much as 267% more for a single month than it did five years ago in areas located near significant data center activity.”8 Additionally, CNN released a report highlighting the personal experience of Lindsey Martin, a Kentucky nurse who lives near a data center, who saw the cost of her electricity bill jump from $150 two or three years ago to $314 in July 2025 and go even higher in August 2025 to $372.9 An Energy Information Administration analysis found similar results, stating that the average price of residential electricity in America has “increased 13% since 2022.”10
President Donald Trump signed an executive order titled “Accelerating Federal Permitting of Data Center Infrastructure” on July 23, 2025, which established a policy to streamline and hasten the development of large-scale data centers—especially those related to AI—by reducing regulatory burdens and utilizing available federal resources.11 In practice, data centers are not just here to stay but to proliferate. And with this reality comes the public policy question: What, if anything, should be done?
Arguments in Favor of Requiring AI Companies to Pay
Ari Peskoe, director of the Electricity Law Initiative at Harvard Law School, thinks these recent AI advances, and their subsequent energy demands, require a reevaluation of the fundamental assumptions of utility regulation, because “when it’s a single consumer that is using so much energy—basically that of an entire city”—the current approach of spreading costs is unfair.12 Former Deputy Undersecretary of Energy David M. Klaus and Georgetown University Adjunct Professor Mark MacCarthy agree with Peskoe’s sentiments and argue, “Data center developers that require massive amounts of electric power … should be required to pay for building new generating and transmission facilities.”13
Supporters of this type of policy point to a July 2025 ruling by the Public Utilities Commission of Ohio as a promising first step and powerful precedent for how to deal with this current technological reality.14 In the decision, the Commission approved a decree by AEP Ohio (the state’s public utility company) to levy special tariffs against tech companies that operate Ohio-based data centers (including Amazon, Facebook, Google, and Microsoft) to pay for the enormous amounts of electricity and all the upgrades and investments to the electrical grid that their data centers require.15 The ruling was celebrated by the Ohio Consumers’ Counsel as one that “protect[s] consumers … from bearing unfair costs.”16 At least 36 utilities across the country have since adopted similar large-load tariffs, with dozens more pending before state regulators.
Proponents of mandatory tariffs argue that voluntary commitments from tech companies are not enough. They point out that even as companies make pledges, independent analyses continue to show electricity costs rising sharply in communities near data centers—and that without enforceable requirements, there is no guarantee those pledges will hold.
Arguments Against Requiring AI Companies to Pay
On the other hand, the tech companies themselves argue that the industry is already moving to address the problem and that mandatory tariffs risk discouraging the investment and innovation that data centers bring to local communities. They contend that data centers pour hundreds of millions of dollars into local communities all across the United States, bring new jobs to their respective regions, and initiate grid upgrades that ultimately help local businesses and improve residents’ quality of life.17
Amazon has pushed back directly on the claim that data centers raise residential electricity bills. In a white paper released in December 2025 in response to a Senate inquiry, the company argued that its data centers “are not being subsidized by other utility customers” and that in some regions, they “have generated surplus utility revenues” that create “downward pressure on rates.”18 Amazon’s research projects that its data centers generate tens of thousands of dollars per megawatt in surplus value for utilities—revenue that can be used to modernize grid infrastructure for the benefit of all customers.
Microsoft has gone further, committing publicly to paying its own electricity costs rather than passing them to ratepayers. Microsoft President Brad Smith wrote in a January 2026 blog post: “Especially when tech companies are so profitable, we believe that it’s both unfair and politically unrealistic for our industry to ask the public to shoulder added electricity costs for AI. Instead, we believe the long-term success of AI infrastructure requires that tech companies pay their own way for the electricity costs they create.”19 Microsoft announced it would work with utilities and state public utility commissions to set tariff rates high enough to cover the infrastructure its data centers require—an approach it says it has already implemented in communities in Wyoming and Wisconsin.
Additionally, tech companies see their AI models and associated data centers as having broader utility that extends to all Americans. They assert that “large-scale data centers spur advancements in AI applications, machine learning, data analytics, and robotics, which in turn … lead to more efficient manufacturing, autonomous transportation systems, the development of new drugs, [and] personalized learning.” They argue that these benefits apply to consumers, businesses, governmental actors, and health and educational institutions from all across the United States—not just those near data centers.20 For the tech companies, the future prosperity of the entire country is dependent on data centers, and voluntary action by the industry is preferable to regulation.
Critics of the voluntary approach, however, note that Microsoft’s pledge came just one day after President Trump posted on Truth Social that tech companies “must pay their own way”—raising questions about whether the commitment reflects a genuine shift or a response to political pressure. A CNN analysis concluded that “there’s not much to enforce” the companies’ promises.21
Discussion Questions
- How often do you use AI in your daily life? Before reading, did you know of the effects it can have on a community’s electricity costs?
- What are the benefits and drawbacks of the current policy of the utility industry?
- Whose arguments do you find most convincing: those advocating for mandatory requirements on AI companies, or those arguing that the industry’s voluntary commitments are sufficient? Why?
- The Public Utilities Commission of Ohio ruled that tech companies must pay special tariffs for their data centers’ electricity use. Should other states follow Ohio’s lead, or could such requirements discourage economic investment?
- Microsoft pledged to “pay its own way” for data center electricity costs one day after President Trump called on tech companies to do so. Does it matter whether a company acts voluntarily or is compelled by regulation, as long as the outcome is the same?
- How, if at all, should local, state, or federal government get involved in regulating data center electricity costs?
As always, we encourage you to join the discussion with your comments or questions below.
Close Up is proud to be the nation’s leading nonprofit civic education organization, working with schools and districts across the country since 1971. If you would like to partner with us or learn more about our experiential learning programs, professional development, or curriculum design and consulting, contact us today!
Sources
Image Credit: alexgo.photography / Shutterstock.com
[1] Harvard Law Today: https://hls.harvard.edu/today/how-data-centers-may-lead-to-higher-electricity-bills/
[2] International Energy Agency: https://www.iea.org/reports/energy-and-ai/energy-demand-from-ai
[3] Union of Concerned Scientists: https://blog.ucs.org/mike-jacobs/data-centers-are-already-increasing-your-energy-bills/
[4] U.S. Department of Energy: https://www.energy.gov/articles/doe-releases-new-report-evaluating-increase-electricity-demand-data-centers
[5] New York Times: https://www.nytimes.com/2025/08/14/business/energy-environment/ai-data-centers-electricity-costs.html
[6] PBS NewsHour: https://www.pbs.org/newshour/show/how-ai-infrastructure-is-driving-a-sharp-rise-in-electricity-bills
[7] New York Times: https://www.nytimes.com/2025/08/14/business/energy-environment/ai-data-centers-electricity-costs.html
[8] Bloomberg: https://www.bloomberg.com/graphics/2025-ai-data-centers-electricity-prices/
[9] CNN: https://www.cnn.com/2025/10/17/tech/electricity-bill-price-increase-ai-data-centers
[10] Energy Information Administration: https://www.eia.gov/todayinenergy/detail.php?id=65284
[11] The White House: https://www.whitehouse.gov/presidential-actions/2025/07/accelerating-federal-permitting-of-data-center-infrastructure/
[12] Harvard Law Today: https://hls.harvard.edu/today/how-data-centers-may-lead-to-higher-electricity-bills/
[13] The Hill: https://thehill.com/opinion/energy-environment/5420210-data-centers-power-demand-ai/
[14] Signal Ohio: https://signalohio.org/data-centers-not-ohioans-must-pay-costs-of-new-power-grid-investments/
[15] Public Utilities Commission of Ohio: https://dis.puc.state.oh.us/ViewImage.aspx?CMID=A1001001A25G09B43531I00509
[16] Office of the Ohio Consumers’ Counsel: https://www.occ.ohio.gov/content/data-center-costs-24-0508-el-ata
[17] New York Times: https://www.nytimes.com/2025/08/14/business/energy-environment/ai-data-centers-electricity-costs.html
[18] GeekWire: https://www.geekwire.com/2025/senators-probe-ai-tech-giants-over-electric-bills-amazon-says-its-data-centers-pay-more-than-their-share/
[19] Microsoft via Latitude Media: https://www.latitudemedia.com/news/microsoft-to-pay-its-own-way-for-power-hungry-data-centers/
[20] Brookings Institution: https://www.brookings.edu/articles/the-future-of-data-centers/
[21] CNN: https://www.cnn.com/2026/01/22/climate/big-tech-warren-electricity-data-centers
