Federal energy dominance meets state utility regulatory authority — and neither side is backing down
If you have spent more than a few years in the electric power business, you have operated under a clear division of authority. The Federal Energy Regulatory Commission oversees interstate transmission and wholesale electricity markets, while state public utility commissions regulate retail rates, local distribution infrastructure, and most generation and transmission siting decisions. That split, codified in the Federal Power Act of 1935, has governed how utilities plan capital spending, structure tariffs, and negotiate with regulators for nearly 90 years.
That division is now being redrawn in real time, and the new version is messier, more contested, and far less predictable than anything the law’s drafters anticipated.
The trigger is data centers. The pace and scale of AI-driven electricity demand have exposed a gap between where federal authority ends and state authority begins, a gap the current regulatory framework was never designed to handle. The Trump administration’s energy-dominance agenda has turned that gap into a genuine fault line, and utility managers from Virginia to California are now making multibillion-dollar investment decisions in the middle of an active federal-state authority battle. Understanding what is actually happening, not just the headlines, but the specific rulings and rulemaking proceedings driving the conflict, is now basic operating knowledge for anyone in a management seat.
The Line That Used to Hold
Section 201(b)(1) of the Federal Power Act gives FERC authority over interstate electricity transmission and wholesale power sales, while preserving state authority over retail sales, local distribution facilities, and generation facilities except where the statute specifically provides otherwise. That is the framework.
In practice, that line has always bent a little. FERC’s Order 1000 expanded federal requirements around regional transmission planning and cost allocation. Order 2023 reformed generator interconnection procedures across transmission providers. Order 1920, the most recent landmark, mandated long-term regional transmission planning using at least a 20-year planning horizon. Each order pushed the federal footprint further into territory that states once managed on their own.
What nobody fully anticipated was a class of electricity customer that would arrive quickly enough, and at sufficient scale, to strain the framework. An AI data center drawing hundreds of megawatts is not like a large manufacturer or a university campus. It may seek transmission-level service, co-located supply, or other arrangements that sit close to the boundary between federal transmission authority and state retail regulation. That puts it squarely in regulatory no-man’s-land: the transmission connection may fall within FERC’s domain, while the retail sale of electricity to that customer generally remains a state matter. That no-man’s-land is exactly where the current fight is concentrated.
Enter the National Energy Dominance Council
On February 14, 2025, President Trump signed an executive order establishing the National Energy Dominance Council within the Executive Office of the President, chaired by Interior Secretary Doug Burgum and vice-chaired by Energy Secretary Chris Wright. Its mandate was to advise the President on a national energy-dominance strategy by identifying ways to improve permitting, production, generation, distribution, regulation, and transportation across U.S. energy sectors.
The Council quickly became part of the administration’s broader energy-policy push. By early 2026, FERC was citing initiatives involving the National Energy Dominance Council, the January 2026 PJM Statement of Principles, and the President’s Ratepayer Protection Pledge as part of the policy backdrop for its large-load work. That is more precise than saying the Council itself directly controlled PJM market rules, but the practical effect for utilities was similar: federal energy policy was moving faster and more visibly into wholesale-market and transmission-planning questions.
Separately, on October 23, 2025, Energy Secretary Wright invoked Section 403 of the Department of Energy Organization Act to direct FERC to consider an advance notice of proposed rulemaking on how large electrical loads, generally defined as demand exceeding 20 megawatts, interconnect with the interstate transmission system. The proposal covered large-load customers, including data centers and manufacturing facilities, and asked whether reforms were needed to make those interconnections timely, orderly, reliable, and non-discriminatory. Because load interconnection had historically been handled largely through state-regulated retail-service and utility-tariff processes, Wright’s directive immediately drew legal-authority objections from state regulators.
The administration’s position, stated plainly: state regulatory frameworks were not built for AI-scale electricity demand, and the federal government would not wait for them to catch up.
The Amazon-Talen Case That Brought It into Focus
To understand why this is now a boardroom conversation, not just a regulatory law debate, look at what happened with Amazon Web Services and Talen Energy’s Susquehanna nuclear power plant in rural Pennsylvania.
In 2024, PJM filed an amended interconnection service agreement that would have allowed expanded power delivery from Talen’s Susquehanna nuclear facility to an AWS data center campus located next to the plant. The proposal would have increased the co-located load from 300 megawatts to 480 megawatts and raised broader questions about behind-the-meter and co-located service arrangements. FERC rejected the amended agreement 2-1 on November 1, 2024, finding that PJM had not shown the proposed nonconforming provisions were necessary because of specific reliability concerns, novel legal issues, or other unique factors. FERC later denied rehearing in April 2025.
The effects spread quickly. Major technology companies had been pursuing similar co-location arrangements with nuclear plant owners, drawn by the need for 24/7 carbon-free power to run AI workloads. The rejection signaled that FERC would scrutinize every deal on its individual merits, and that physical proximity to a power source was not, by itself, sufficient justification.
More broadly, the case crystallized the control question for utility executives: when a large customer is served through a transmission-connected, co-located arrangement, which parts of the arrangement are federally regulated and which remain subject to state retail authority? FERC’s answer was not a sweeping rule for all future data-center deals. It was narrower and more procedural: PJM had not justified the deviations from its standard agreement in that particular filing. Even so, the decision became a warning sign that co-located load arrangements would face close review at both the federal and state levels.
That is what made the June 2026 orders so significant. FERC was no longer dealing only with one disputed co-location arrangement in Pennsylvania; it was signaling that the underlying problem, how large, transmission-connected loads should be studied, charged, and integrated, had become a regional market issue across the organized grid.
FERC’s June 2026 Orders and What They Actually Say
On June 18, 2026, FERC issued tailored show-cause orders under Section 206 of the Federal Power Act to the six FERC-regulated RTOs and ISOs-PJM, MISO, SPP, CAISO, ISO New England, and NYISO-together with relevant transmission owners. The agency preliminarily found that existing tariff provisions for large and co-located loads may be unjust and unreasonable, and directed each region to either justify its current tariff or propose revisions.
The approach was deliberate. Rather than issuing a single national rule, FERC used region-specific Section 206 orders directed at the organized markets where it regulates transmission service. That choice moved the process faster than a conventional rulemaking and reduced, though did not eliminate, the litigation risk that would come with a broader national standard.
NARUC had filed comments before the orders landed, arguing that end-use sales, retail-service decisions, and many load-interconnection issues remain within state authority under the Federal Power Act regardless of load size. NARUC urged FERC to preserve and affirm state retail regulatory authority while still allowing coordination on timely and non-discriminatory large-load interconnection standards.
FERC threaded the needle by focusing on tariff provisions governing transmission access, interconnection study processes, co-located load treatment, cost transparency, and flexible large-load service, not on state retail ratemaking itself. The American Bar Association described the broader issue as “the jurisdictional collision over large loads,” and that collision is far from over. Grid operators now face response and compliance deadlines, followed by comments, potential rehearing requests, and implementation work that could stretch into 2027.
States Push Back — and Some Make Their Own Moves
State regulators did not stand down. More than 300 data center energy bills were introduced across more than 30 states in just the first six weeks of 2026, ranging from cost-allocation requirements to construction moratoriums in states that wanted time to study the impacts before approving additional data center development.
Texas moved through Senate Bill 6 and related Public Utility Commission of Texas proceedings to address large-load interconnections in ERCOT, including proposed standards for new or expanded loads of 75 megawatts or more, financial-security requirements, and stranded-cost protections. Pennsylvania’s PUC finalized a large-load model tariff framework in May 2026, applying guidance to customers exceeding 50 megawatts individually or 100 megawatts in the aggregate and emphasizing cost-causation and ratepayer protection. New York, meanwhile, opened a PSC proceeding under Governor Hochul’s Energize NY Development initiative to review interconnection processes, cost-allocation mechanisms, and tariff structures for large loads.
The politics here are not cleanly partisan. Pennsylvania Governor Josh Shapiro, a Democrat, secured federal support in 2026 to extend the PJM price cap, aligning with the Trump administration’s affordability goals while protecting Pennsylvania ratepayers. The state coalition challenging FERC’s Order 1920 transmission planning rules was led by 19 Republican attorneys general; 12 Democratic-led states filed in support. On energy federalism, the usual political alignment breaks down, making it harder to forecast where support for any given regulatory position will land.
What This Means for Your Planning Horizon
Here is the part that matters if you manage capital-expenditure decisions, regulatory strategy, or large-customer interconnections at a utility.
The timeline for resolving large-load interconnection disputes has lengthened and become less predictable. FERC’s June orders require the affected RTOs and ISOs to justify existing tariff provisions or propose revisions, followed by comment, rehearing, and implementation steps that could carry disputes into 2027. Any capital plan that assumes near-term regulatory clarity on data center interconnections needs to account for that delay explicitly rather than treating it as a scheduling risk that will sort itself out.
Cost allocation is being decided inconsistently across states. If your service territory requires data centers to bear the full cost of network upgrades they trigger, your interconnection economics look fundamentally different than they do in a state that socializes those costs across the general rate base. Those different answers affect where data centers locate, which in turn shapes your load forecast, generation planning, and transmission investment case. Getting this wrong by one large customer can materially shift a five-year capital plan.
The administration’s broader energy-dominance agenda also changes the risk model for utility executives who built strategic assumptions around a predictable notice-and-comment regulatory cycle. DOE’s Section 403 directive, FERC’s accelerated response, and federal ratepayer-protection messaging show that federal action can move faster than the traditional utility-planning cycle—and it can reach into wholesale-market and transmission-service mechanics that state regulators do not control. Treating federal intervention as a tail risk, rather than a live planning variable, is no longer defensible.
The Federal Power Act is 90 years old. It was designed for a grid that looked nothing like today’s, serving load types that did not exist when the statute was drafted. The stress fractures were always there. AI-scale data center demand just found them all at once.
Conclusion
The conflict between federal energy dominance and state utility regulatory authority is not a problem that one ruling or one piece of legislation will settle. The Federal Power Act’s federal-state boundaries were porous long before the data center boom arrived. The Trump administration’s NEDC, FERC’s June 2026 orders to regional grid operators, and NARUC’s pushback have compressed what might have been a gradual shift in authority into an 18-month sprint, and that sprint is still running.
The practical risk for utility managers is not that this conflict ends badly. It is that it does not end cleanly, or on any timeline that helps build a reliable five-year capital plan. Overlapping federal and state requirements, inconsistent cost-allocation answers across regions, and the possibility of further executive-level intervention in wholesale markets are now standing features of the operating environment, not temporary disruptions.
The managers who navigate this best will be the ones mapping their specific regulatory exposure in each state where they operate, stress-testing capital plans against multiple regulatory outcomes, and maintaining active relationships with both their state PUCs and federal regulators. In a regulatory environment this fluid, the worst position to be in is surprised.
The line between federal and state electricity authority still exists. It just no longer runs where utilities were trained to expect it.