As the 2026 midterm race heats up, members of Congress are racing to prove they understand voter anxiety about the surge of AI data centers and the ripple effect on residential electricity bills. The core of the dispute is simple: who should absorb the extra grid expenses created by megawatt-hungry facilities? The answer, however, proves elusive, sparking a legislative tug-of-war that mirrors the technical complexity of the power market itself.
Competing bills aim to shift costs onto large-load customers
Republicans have introduced the Ratepayer Protection Act championed by Ohio Senator Jon Husted. The proposal would require state regulators to design rates that force data centers consuming more than 100 MW to cover the additional transmission and distribution upgrades they trigger. Husted, who is locked in a fierce re-election battle against former Senator Sherrod Brown, sees the bill as a political shield against criticism that he previously supported data-center expansion in Ohio. The measure cleared the House in September with bipartisan backing, but it stalled in the Senate when Minority Leader Chuck Schumer, a Democrat, objected to its voluntary language.
Schumer instead backs the GRID Savings Act authored by New Mexico Democrat Martin Heinrich. While similar in intent, the GRID Savings Act includes stronger enforcement mechanisms, according to its sponsors, and is positioned as the more robust alternative.
How much are residential customers actually paying?
Experts agree that the precise share of grid costs passed to homeowners varies by utility, state, and contract structure, making a single national figure impossible. In Maryland, for instance, analysis from the University of Maryland’s School of Public Policy suggests that the average household’s electricity bill may be $168 to $216 higher each year solely because of the data-center boom. Connor Waldoch, co-founder and chief strategy officer at Grid Status emphasizes that “tens to hundreds of millions of dollars of infrastructure” are being built by utilities and ultimately reflected on consumer statements.
Lucy Qiu, a professor of public policy, notes that the pass-through depends on power-purchasing agreements, retail rate design, and regulator decisions about cost allocation. When utilities sign long-term contracts that lock in lower wholesale prices, the incremental cost of new demand can be spread across all ratepayers, whereas a shift to spot-market pricing could concentrate the burden on newer, high-consumption loads.
Regional impact: PJM’s $29 billion tally
To grasp the scale, consider the PJM Interconnection, the regional transmission organization that runs the grid for 13 states and the District of Columbia, including the data-center hub of Northern Virginia. PJM’s Independent Market Monitor estimates that data centers have imposed roughly $29 billion in costs on its 67 million ratepayers over the past two years. Ari Peskoe, director of the Harvard Electricity Law Initiative, stresses that these expenses are “spread across the region to every business and resident that has an electricity meter.”
The heavy concentration of facilities in places like Loudoun County, Virginia—home to about 250 data centers—has prompted local utilities such as Dominion Energy to argue that all distribution and transmission upgrades are recovered from customers under the oversight of the Virginia State Corporation Commission.
Secrecy, NDAs and community pushback
Local opposition often centers on the opacity of data-center deals. A University of Mary Washington study found that non-disclosure agreements (NDAs) were present in 80 % of Virginia municipalities where data centers were approved, proposed, or already operating. Residents in Menomonie, Wisconsin, for example, displayed a protest sign on September 24, 2026, demanding transparency about how new facilities would affect their electricity costs.
Dominion Energy’s spokesperson, Jeremy L. Slayton, reiterated that any cost recovery is vetted by the Virginia State Corporation Commission, but critics argue that the NDAs prevent communities from assessing the true financial impact.
Looking ahead: grid upgrades and broader implications
Beyond the immediate cost debate, the AI data-center surge could accelerate long-needed upgrades to the U.S. electricity system. Qiu points out that some new substations and transmission lines were already required to replace weather-damaged assets, while others are essential for growing demand from electric vehicles and other electrified technologies. In this sense, the data-center boom may inadvertently bolster grid resilience, offering “more robust infrastructure” that reduces outage frequency for nearby customers.
Nevertheless, the rapid pace of construction forces utilities to consider less efficient, faster-to-build power plants, a factor that could further lift rates over time. As 70 % of Americans express concern that data centers will raise their electric bills, policymakers must balance the economic benefits of AI-driven computation against the equitable distribution of the associated infrastructure costs.
