At the heart of the legislation is a straightforward question: When a multibillion-dollar technology company requires enormous new investments in the electrical grid, should existing utility customers be responsible for paying those costs?
Supporters of the bill say the answer should be no.
The legislation establishes a federal framework under which state utility regulators would consider special standards for enormous electricity users. The measure focuses on “large-load customers,” including data centers and other information-technology facilities capable of drawing at least 100 megawatts of electricity from a single location or campus.
Rep. Gabe Evans, R-Colo., introduced the House legislation with bipartisan support. Evans has argued that families and small businesses should not be forced to subsidize infrastructure built specifically to accommodate enormous new electricity demands.
“Large load data centers must cover the full costs of any system updates they require, not families or small businesses,” Evans said while promoting the legislation.
That issue becomes particularly important when a new data center connects to an existing utility system.
A hyperscale facility can require far more than simply running additional power through existing lines. Utilities may need to construct or expand generating capacity, transmission infrastructure, substations and distribution equipment before the facility can receive the electricity it needs.
Without specific safeguards, opponents of cost shifting warn that some of those investments could eventually become part of a utility’s broader rate base. In that situation, existing residential and commercial customers could potentially absorb some of the expense through their electricity bills.
The Ratepayer Protection Act seeks to address that concern by requiring regulators to consider rate structures designed to recover the “full, incremental cost” of infrastructure upgrades needed to serve covered large-load customers.
The proposal also addresses what happens when a proposed data center never materializes.
Utilities can face significant costs when they begin preparing for a massive new customer. If the project is later canceled, delayed or substantially reduced, the utility could be left with infrastructure expenses that were incurred based on electricity demand that never arrives.
The legislation therefore calls for consideration of financial assurances from large customers before major infrastructure investments are made. Such guarantees could help protect utilities and existing customers if a data-center project fails to move forward or terminates its power agreement prematurely.
Importantly, the bill does not completely federalize electricity regulation.
Instead, it amends the Public Utility Regulatory Policies Act of 1978, commonly known as PURPA, by establishing a federal “must consider” standard. State regulators would be required to consider the principles outlined in the legislation while retaining substantial authority over their own utility markets and rate structures.
That approach has produced some disagreement even among groups that support protecting existing ratepayers.
The American Public Power Association has supported efforts to prevent existing customers from being forced to absorb data-center-related costs, while also expressing concerns about imposing another federal PURPA requirement on public power utilities that already have large-load policies.
Another criticism is that the bill’s approach focuses specifically on data centers. The Center for Data Innovation argued that other electricity-intensive industries, such as semiconductor manufacturing, steel production and chemical processing, can create similar grid demands and therefore could also warrant consideration under the same standard.
Despite those disagreements, the legislation attracted overwhelming support in the House.
The Energy and Commerce Committee approved the measure unanimously, 52-0, in July. The full House later approved it 417-3, demonstrating unusually broad bipartisan support for addressing the financial impact of the data-center boom.
The debate comes as America’s electricity demand is accelerating after years of comparatively modest growth.
The U.S. Energy Information Administration now expects American electricity sales to reach a record 4,135 billion kilowatt-hours in 2026, nearly 2% above 2025. EIA projects another increase of roughly 2% in 2027, bringing total electricity sales to approximately 4,211 billion kilowatt-hours.
Commercial electricity demand is increasing even faster. EIA forecasts commercial-sector electricity sales will rise 3.3% in 2026 and another 2.7% in 2027, with data-center development and manufacturing activity among the major drivers of rising demand.
The long-term implications could be even larger.
EIA estimates that data-center servers accounted for approximately 7% of commercial-sector electricity consumption in 2025. Under its projections, server electricity use could represent between 22% and 33% of commercial building electricity consumption by 2050, depending on how quickly computing demand and data-center capacity expand.
Meanwhile, utilities are dealing with another complication: not every proposed data center may actually get built.
Reuters reported earlier this month that data-center electricity requests across the country had surpassed 700 gigawatts, more than ten times the estimated electricity consumption of existing U.S. data centers. The phenomenon has raised concerns that some connection requests may be speculative rather than firm commitments.
Texas has become one of the most prominent examples of the problem. Gov. Greg Abbott recently ordered a halt to new state-issued data-center permits while officials conduct an audit examining the industry’s impact on the electrical grid. Reuters reported that proposed Texas data-center projects seeking grid connections totaled more than 470 gigawatts — several times the state’s peak electricity demand.
The issue therefore extends beyond a single piece of legislation.
America is racing to expand artificial-intelligence infrastructure while simultaneously trying to build enough power generation and grid capacity to support it. The central policy dispute is increasingly about how those costs should be allocated.
For supporters of the Ratepayer Protection Act, the legislation provides a mechanism to make companies driving extraordinary new electricity demand responsible for the infrastructure required to serve them.
For critics, the question is whether a federal “must consider” requirement is necessary when states and utilities are already developing their own policies.
For now, the House has spoken with an overwhelming bipartisan vote. But after the Senate objection, the legislation’s future remains unsettled — leaving one of the biggest questions surrounding America’s AI expansion still on the table: who ultimately pays for the power needed to fuel it?
