Utilities, grid operators, and large customers are converging on a new question: who should carry the cost of power for hyperscale growth?

FirstEnergy has opened a new debate over how the PJM Interconnection should handle the rapid rise of data center electricity demand. During the company’s quarterly earnings call, Chief Executive Brian Tierney said PJM’s proposed Reliability Backstop Procurement auction could be a step in the right direction, but objected to key parts of the design, especially the grid operator’s role as an intermediary between power plant developers and end-use customers.

The argument is more than market design detail. It goes to the center of one of the most consequential utility questions now facing the eastern United States: how to connect and serve a wave of large data center projects without forcing traditional customers to absorb costs created by a narrow group of fast-growing, power-hungry users. As artificial intelligence infrastructure expands and developers race to secure capacity, regulators and utilities are being pressed to decide who signs the contracts, who bears the risk, and who ultimately pays.

FirstEnergy wants direct contracting, not another layer in the middle

Tierney’s criticism was specific. In his view, the parties making the investment in new generation should be the developers and builders of the power plants contracting directly with the end-use customers that need the electricity. He argued that inserting PJM into the middle of that arrangement, with electric distribution companies also acting as intermediaries, creates a structure in which the wrong people may end up paying.

That is a notable stance because it frames the data center buildout as something that should be financed as close to the source of demand as possible. Rather than socializing costs broadly across the system, FirstEnergy is signaling that large-load customers should have a clearer and more direct commercial responsibility for the infrastructure their projects require. In a region where affordability has become politically sensitive, that message is likely to resonate well beyond one earnings call.

PJM’s backstop concept reflects a real system problem. New load is arriving at a scale and speed that can outpace traditional planning and procurement cycles. Grid operators want mechanisms that prevent reliability gaps when generation or network upgrades lag demand growth. But once a centralized backstop exists, the question becomes whether it functions as a neutral reliability tool or as a channel through which costs can migrate from specialized customers to the broader rate base.

The scale of the queue shows why the dispute matters

FirstEnergy’s own numbers show why the issue has become urgent. The company said it has 4.3 gigawatts of contracted data centers expected to come online by 2031, nearly 50% higher than the level it reported in February 2025. Beyond those projects, its utilities have a pipeline of potential data center developments from what it described as reputable customers that meet certain project metrics. That pipeline has grown to 7.4 gigawatts by 2031 and 14.9 gigawatts by 2035, an increase of about 15% since February.

Those figures are striking not only because they are large, but because they suggest the market is still accelerating. A pipeline measured in the tens of gigawatts is not a marginal planning issue. It implies lasting consequences for generation investment, transmission expansion, interconnection timelines, and retail rate debates across multiple states.

FirstEnergy serves about 6 million customers in Ohio, Pennsylvania, New Jersey, West Virginia, Maryland, and New York, mainly within PJM. That footprint means its view carries weight in the region’s evolving argument over load growth and cost allocation. If utilities believe existing customers could end up subsidizing data center expansion, pressure will build for tougher tariffs, more direct large-load contracting models, or stricter standards for proving project readiness before grid capacity is reserved.

The affordability and political backdrop is already tightening

The company’s pushback comes at a moment when affordability concerns are becoming harder to separate from reliability planning. According to the article, Pennsylvania Gov. Josh Shapiro said his administration will oppose rate hike requests that fail to meet legal requirements and address affordability. That is a reminder that any market design perceived to shift costs onto households and small businesses will face scrutiny from state officials, even if the original proposal is rooted in reliability concerns.

For utilities, this creates a balancing act. They want to capture the upside of major load additions, including stronger sales growth and justification for network investment. FirstEnergy’s first-quarter results point to that opportunity. The company reported $405 million in attributable earnings, up 12.5% from a year earlier, and revenue rising to $4.2 billion from $3.8 billion. But the same growth story also raises the stakes of getting planning rules wrong. If the economics of the data center boom become politically associated with higher bills for ordinary consumers, the sector could face a backlash.

That helps explain Tierney’s insistence on aligning investment responsibility with the customer driving the demand. It is not only a commercial preference. It is also a defensive regulatory position, one that tries to preserve support for grid expansion by reducing the appearance of cross-subsidy.

What comes next for PJM and the broader market

The coming fight will likely center on whether PJM can design a backstop mechanism that preserves reliability without blurring accountability. Large customers want speed and certainty. Utilities want clarity on who pays. Regulators want to avoid bill shock. Grid operators want a workable system before demand outruns supply additions. Those priorities overlap, but not perfectly.

FirstEnergy’s intervention suggests the next phase of the data center power buildout will be shaped as much by market architecture as by physical infrastructure. Transmission lines, substations, and gas turbines may be the visible pieces, but the harder question is contractual: who is on the hook if projected demand arrives late, costs rise, or procurement decisions miss the mark?

As PJM refines its proposal, the region is effectively testing a broader national model. If direct contracting becomes the preferred answer, other markets may move the same way. If centralized backstops prevail, states and utilities will keep pressing for stronger protections against broad cost transfer. Either way, the era when explosive data center growth could be treated as just another source of load is ending. It is now a structural question for the grid itself.

This article is based on reporting by Utility Dive. Read the original article.