> ## Content Index > Fetch the complete content index at: https://www.growthsolutionskc.com/llms.txt > Use this file to discover other available public pages before exploring further. # Who Pays for the AI Power Boom? - URL: https://www.growthsolutionskc.com/who-pays-for-the-ai-power-boom/ - Published: 2026-09-21T14:00:07.000Z - Updated: 2026-09-21T14:00:06.000Z - Author: Matt Cucinotta - Tags: Tech & Innovation, #growthsolutionskc, #series-AI-energy ## Why Data Centers Do Not Automatically Mean Higher Electric Bills *Part 1 of the Growth Solutions KC AI Energy Series* --- The headlines are easy to understand. AI data centers use enormous amounts of electricity. Utilities need new power plants, transmission lines, substations, and transformers. Someone has to pay for all of it. So the conclusion can feel obvious: Your electric bill is going up because Big Tech is building data centers. Sometimes new data-center demand can contribute to higher costs. But that simple narrative skips the most important part of the story. Who actually pays depends on how the project is connected, what new infrastructure it requires, how regulators allocate those costs, what contracts the developer signs, whether the project ever gets built, and whether new infrastructure also benefits the broader grid. The question is not whether AI requires more electricity. It does. The better question is: **Who should pay for the infrastructure required to deliver it?** That distinction matters because America can build the power system needed for the AI era without automatically handing the bill to households. --- ## What Matters: The Load Is Real, but the Bill Is Not Automatic The scale of electricity demand is changing quickly. Data centers consumed about 4.4% of U.S. electricity in 2023\. Updated Lawrence Berkeley National Laboratory scenarios cited by the Department of Energy estimate they could account for roughly 9.5% to 15.3% by 2030. That is a serious infrastructure challenge. A large AI campus can require new generation, transmission capacity, substations, transformers, and local distribution equipment. Those assets require major capital investment and can take years to plan and build. But a utility bill is not a direct meter attached to national data-center electricity consumption. Electricity rates are shaped by regional power markets, utility regulation, fuel costs, existing infrastructure, new capital investment, customer growth, and—critically—the rules governing who pays for specific upgrades. That means two communities can experience similar data-center growth and very different outcomes. The key is cost allocation. --- ## The Driveway and the Highway ![](https://storage.ghost.io/c/42/b4/42b4b6b8-b1f2-4aa3-8c91-88f72e71a120/content/images/2026/09/Driveway-to-the-highway.png) A useful way to think about the problem is the difference between a driveway and a highway. The driveway is the infrastructure built specifically to connect one large customer: the substation at the site, dedicated wires, and other equipment needed to bring electricity directly to the data center. The logic here is relatively straightforward. If a private development needs a private driveway, the developer should pay for it. **The highway is harder.** A massive new load may also require upgrades miles away: larger transmission lines, new generation, system reinforcement, or equipment that becomes part of the shared electric grid. The research dossier identifies this distinction as one of the central questions in the emerging cost-allocation debate. Now the question changes. If the upgrade exists only because one data center arrived, should every customer help pay for it? - What if the upgrade also improves reliability for thousands of homes and businesses? - What if several new factories, neighborhoods, or future data centers eventually use the same capacity? That is where simple slogans stop being useful. Good regulation has to distinguish between infrastructure created primarily for one customer and infrastructure that produces broader public benefits. --- ## The Risk of Cost Shifting The concern about ratepayers is legitimate. If a utility spends heavily to serve a new data center and then spreads those costs across its entire customer base, households and small businesses can end up financing infrastructure created mainly for a very large commercial user. There is an additional problem: What happens if the project never arrives? Developers searching for the fastest path to power can request capacity from multiple utilities or locations while deciding where to build. FERC has warned that speculative or duplicative requests can clog planning processes, distort load forecasts, and contribute to unnecessary costs. Imagine a utility building a six-lane road for traffic that never comes. Someone still has to pay for the road. That is the stranded-asset problem. And it is why the answer cannot simply be “build more.” Utilities and regulators also need credible commitments from the customers asking them to build. ![](https://storage.ghost.io/c/42/b4/42b4b6b8-b1f2-4aa3-8c91-88f72e71a120/content/images/2026/09/Risk-of-cost-shifting.png) ### The Rules Are Beginning to Change This is already becoming a major regulatory issue. In June 2026, the Federal Energy Regulatory Commission ordered all six regional grid operators under its jurisdiction to justify or reform how they handle large electricity users. Among the issues FERC specifically identified were preventing cost shifting, increasing transparency, accommodating behind-the-meter generation, creating flexible service options, and improving the study process for large loads. One of the emerging tools is a cost-recovery agreement. The basic idea is simple: if infrastructure is built for a large data-center project, the large customer should remain financially responsible for its fair share even if the project is delayed or never comes online as planned. FERC Commissioner David Rosner described these agreements as a way to prevent residential customers from being left with costs created for projects that fail to materialize. Other approaches can include separate large-load tariffs, minimum billing commitments, upfront deposits, long-term contracts, and requirements that developers fund dedicated network upgrades. The White House's 2026 Ratepayer Protection Pledge pushes in the same direction. Participating companies commit to build, bring, or buy the additional power they require, pay for delivery infrastructure needed to serve their data centers, and negotiate separate rate structures requiring payment for infrastructure brought online on their behalf—even if they ultimately use less electricity than expected. The principle is more important than the politics around any particular program: Growth should pay its own way. ### “Pay Your Own Way” Requires Judgment That principle sounds simple until we reach the highway again. Not every transmission upgrade benefits only a data center. - A new substation may strengthen local reliability. - A transmission project may unlock capacity for future businesses. - New generation can increase regional supply. - Grid-enhancing technologies may allow existing wires to carry more electricity. A flexible data center may even agree to reduce its grid consumption during periods of system stress. FERC is specifically encouraging flexible service and technologies that can reduce the need for expensive new network upgrades. In those cases, assigning 100% of every cost to one customer may not accurately reflect who benefits. The objective should therefore be cost causation, transparency, and durable commitment—not punishment. Questions needing clarity and understanding: - Who caused the need for the project? - Who benefits from it? - Who bears the financial risk? - What happens if expected demand never appears? Those questions are more useful than assuming either that every data center is subsidized or that every grid investment automatically benefits everyone. ### Can Data Centers Help Lower Costs? Potentially. But there are no guarantees. If very large electricity customers finance new generation and transmission, contribute fixed payments over long periods, or expand the customer base without shifting their infrastructure costs to existing users, the result can strengthen the broader system. New infrastructure can also create capacity or reliability benefits that later serve other customers. DOE is already highlighting projects where data-center developers are funding generation, transmission, and other infrastructure rather than placing those costs on households. But the opposite can happen if supply fails to keep up, infrastructure is poorly planned, speculative projects cause overbuilding, or regulators allow costs to migrate onto households. That is why the answer is not ideological. It is structural. The outcome depends on the rules. --- ## What People Should Understand ![](https://storage.ghost.io/c/42/b4/42b4b6b8-b1f2-4aa3-8c91-88f72e71a120/content/images/2026/09/4-questions.png) When someone says: “AI data centers are going to raise your electric bill.” Do not automatically accept the claim. Do not automatically dismiss it either. Ask four questions: - What new infrastructure is actually required? - Who is contractually responsible for paying for it? - What happens if the data center never uses the capacity it requested? - Does the new infrastructure serve only the data center, or does it create measurable benefits for the broader grid? Those questions move the conversation away from fear and toward accountability. Data centers are enormous electricity users. Their growth will require real investment, and poorly designed rules can absolutely shift costs toward people who had no role in creating them. But higher household bills are not an unavoidable law of AI. They are partly a consequence of market design, regulatory decisions, infrastructure planning, and contractual responsibility. --- ## What Happens Next The AI energy buildout is forcing utilities, regulators, technology companies, and communities to reconsider rules designed for a much slower era of electricity growth. That is healthy. America needs more power infrastructure. It also needs rules that distinguish private costs from public benefits, filter speculative projects, protect customers from stranded investments, and allow major new users to connect without weakening reliability or affordability. The goal should not be to stop data-center growth. Nor should the goal be to subsidize it. The goal should be a system in which growth pays for the costs it creates, communities share in genuine benefits, and infrastructure investment strengthens rather than strains the electric system. The next time someone asks who pays for the AI power boom, the answer should not automatically be: You do. The better answer is: It depends on the rules we build around it. **And those rules are being rewritten now.** That leads to the next question in the AI energy story: **If America is going to need far more reliable power, where will it come from?** --- — Matt Cucinotta | Growth Solutions KC | Inspire · Inform · Ignite --- > Source and methodology note: This article draws on the Growth Solutions KC AI Energy Research Dossier and current primary-source material from the U.S. Department of Energy, Federal Energy Regulatory Commission, and federal ratepayer-protection initiatives. Electricity-market structures and utility-rate rules vary substantially by state and region. Specific projects and consumer-rate effects should be evaluated within their relevant regulatory jurisdiction.