Reimagining service obligations in an age of mighty prosumers

Key points

• Large-scale data centers are a distinct type of electricity user, able to generate their own power and move locations as needed. Their rise as proactive energy consumers is changing the landscape of utility markets and prompting reconsideration of service obligations and consumer protection policies.
• States are developing multi-layered frameworks for large loads. Meanwhile, the U.S. Department of Energy is pushing for a national interconnection regime to expedite data center connections, potentially bypassing traditional state regulatory roles. 
• The need for coordinated governance is growing. Electric cooperatives are calling for federal-state collaboration, transparent processes, and updated policies to balance universal service with risk protection — all while ensuring fair cost allocation and infrastructure investment.

There are several reasons electric cooperatives choose to serve large loads in their communities. The most important of which is that the commitment to serve all members is a co-op foundational value. This commitment harkens back to the first “golden age of electrification“ when electricity access for rural communities couldn’t be taken for granted, was often memorialized in a co-op’s mission statement and likely codified in state regulations governing electric utilities. Co-op websites, public-facing documents from the National Rural Electric Cooperative Association and other authoritative sources continue to confirm the centrality of the duty to serve but do not quantify how many cooperatives explicitly include this language in their mission statements. Yet we intuitively recognize that this obligation is a universally shared value.

So, what happens when the commitment to serve potentially imperils a co-op’s financial or operational well-being?

In today’s age of accelerated electricity demand growth, the sheer size and number of interconnection requests from large commercial and industrial (C&I) consumers are unprecedented. Many of these new customers are much larger and better capitalized than the electric cooperatives that serve them. Moreover, they can bring their own generation to the bargaining table in various forms (continuous or back-up and owned or contracted supply) and have ample opportunities to site their facilities wherever they choose. Without a doubt, the rise of these large, sophisticated buyers has decisively shifted the balance of bargaining power in the marketplace, raising questions whether attitudes toward service obligations need to adapt, consumer protection policies need to evolve and possibly, which regulator should be in the driver’s seat for developing and overseeing new industry guardrails.

Is the regulatory compact outdated?

The electricity market has clearly changed since the regulatory compact emerged in the early 1900s, protecting smaller, less organized buyers with less bargaining power and few competitive alternatives against potential bad practices of well-capitalized electric utilities. While the framework has undergone at least one major policy recalibration it is still largely intact in most states, especially for noncompetitive segments of the industry, such as transmission and distribution services. Because the Federal Power Act reserves regulation of retail electricity sales and local distribution to the states, the core responsibilities that affect electricity consumers arise under state regulatory compacts, with the federal government’s role intentionally narrowed to wholesale markets and interstate transmission.

The traditional state regulatory compact is built on a three‑part exchange whereby: the utility is granted a monopoly franchise (exclusive service area) and the right to recover costs plus a reasonable profit in exchange for an obligation to serve everyone reliably and safely at fair rates (creating an inherent, reciprocal consumer obligation to take the utility service and pay the approved rates). This structure is overseen by a state regulator (or, under limited conditions, by a federal regulator) balancing public versus investor interests. Yet, given the accelerated changes that have taken place in the last two decades — with an erosion of traditional utility market power beginning with market restructuring, new regulatory gaps created through deregulation and more expansive consumer capabilities — the compact feels increasingly out of step with the industry’s operational reality.

The policy recalibration that set this change in motion took place in the 1990s and early 2000s, with several states taking direct aim at the “monopoly” structure and the inherent consumer “obligation” by deregulating or restructuring their wholesale and retail electricity markets in response to rising electricity rates. These initiatives significantly shifted market power, creating regulatory gaps and generating foundational cracks in the regulatory compact. The more important changes that have transpired and that take direct aim at traditional integrated utility market power include:

1. Competitive generation markets have allowed independent power producers to displace utility‑owned generation.
2. Retail choice in some states has diminished the concept of a captive customer base by enabling consumers to choose alternative suppliers.
3. Market‑based pricing mechanisms have replaced cost‑of‑service generation — shifting control of dispatch and price formation to Regional Transmission Organizations/Independent System Operators market pools.
4. Bilateral contracting and wholesale market participation further changed the power procurement dynamic in the marketplace.

It is important to understand that retail choice was created to introduce competition, not to extend the traditional utility service mandate to competitive providers. Consequently, competitive suppliers operate under meaningfully lighter regulations than traditional utilities, as they compete mainly on price and the type of energy supply rather than fulfilling universal service obligations. Competitive suppliers typically can choose which customer classes they market to (large C&I vs. residential), which regions they serve and which customers they find economically attractive. In restructured states, utilities must still offer default service for customers who do not choose a competitive supplier as the Provider of Last Resort, reinforcing that retail suppliers do not bear the universal service duty. Moreover, when customers enter third-party energy service contracts, power purchase agreements, or leases, oversight shifts away from public utility commissions — meaning that the regulatory compact no longer covers large, important parts of the electricity ecosystem.

How prosumers, DERs and mobile loads are transforming demand

But it’s not only supply-side competition that has altered our perspective on market power. Important changes in consumer behavior have also redefined how we think about duty and regulatory guardrails. The original regulatory compact assumed consumers were passive load-takers, simply buying utility-provided power at regulated rates. The rise of distributed energy resources (DERs) in the 2010s — such as rooftop solar, home batteries, electric vehicles, and demand-response technologies — transformed consumers into “prosumers” who generate, store and manage their own energy.

Retail choice was created to introduce competition, not to extend the traditional utility service mandate to competitive providers.

Due to high costs, complex permitting and limited policies, DERs were at one time used mainly by large commercial customers, having minimal impact on the overall structure of the grid. Now, with supportive policies, lower technology costs and regulatory changes like FERC Order 2222, consumer-owned generation is widespread. Many electricity consumers can now sell power back to the grid or participate in demand response, becoming active market players alongside utilities.

The recent proliferation of very mobile, large-scale entities such as data centers has notably enhanced the concept of prosumers, fundamentally shifting the balance of power between electricity purchasers and suppliers. The increased hybridization of data center development — with greater dependence on leased colocation facilities — suggests that this mobility or locational flexibility is increasing, with less land and infrastructure exposure for the ultimate electricity consumer. According to recent analysis by McKinsey, 64% of the capital investment made in data center infrastructure between now and 2030 will be in servers and storage, equipment that can be relocated in the event of downsizing or consolidation. The key point being that, from a grid‑planning and regulatory standpoint, large data‑center loads are materially more mobile than traditional utility customers, even if relocation is not instantaneous.

More importantly, these commercial actors possess the capacity to generate utility-scale power independently, establish direct supply agreements, and may potentially operate without reliance on local utility mandates or within the boundaries established by the traditional electricity grid. An article from The Wall Street Journal goes so far as to suggest that the utility model might be obsolete with data center power supply plans now including options to build primary rather than back-up power solutions. One notable example of the go-it-alone approach mentioned in the article was OpenAI’s massive Stargate initiative.

Stargate Project resets market dynamics, resource allocation

The White House’s January 2025 announcement of the Stargate Project, a $500 billion, 10-gigawatt AI infrastructure public-private initiative, established artificial intelligence development as a national priority alongside semiconductors and energy independence. On the second day of his return to office, President Trump emphasized Stargate’s role as a central element of his administration’s strategy to advance American technology, strengthen energy security and maintain economic leadership, emphasizing its significance for modernizing the nation during remarks made aptly in the Roosevelt Room. While the project presents a highly ambitious agenda — one that will continue to face substantial financial, technical, logistical and political challenges — the ongoing efforts to address these obstacles should fundamentally reset our collective perspectives on utility market dynamics and resource allocation.

Twin sites near Abilene, Texas and a third location in Milam County within the state are considered “ground zero” for the Stargate initiative and illustrate how gigawatt-scale development to achieve AI economies of scale might unfold. Texas was intentionally chosen because of its vast amount of undeveloped land suitable for these massive data‑center campuses, ample and sometimes stranded energy resources — think of abundant West Texas wind and the natural gas from Permian oil drilling — and highly competitive state and local incentives.

The electricity consumption of the 1.2 GW Stargate Abilene site rivals that of Denver, Colorado.

The flagship Abilene Stargate site originally had the codename “Project Ludicrous” because of the incredible speed and massive scale at which it is being constructed. In fact, development was already underway before the Stargate initiative was announced, thanks to a creative partnership between Lancium — which supplied the land and power infrastructure — and Crusoe Energy, which builds, owns, and operates the data center facilities on-site. The momentum of the 1.2 GW megaproject made the site a natural anchor for a national program seeking speed and scale.

But it is the unparalleled pace of the project, which has gone from an energy-first concept, planned on a remote, undeveloped site in West Texas to an energized outpost in 12 months as the first of a forming archipelago of massive compute clusters, that signals the arrival of a new kind of electricity consumer. As a CNBC reporter observed, “Meta, Google, xAI, OpenAI, Microsoft and Amazon are planting hyperscale campuses across the heartland, turning farmland and factory shells into compute factories that rival cities in electricity demand.”

These major load consumers possess substantial capital resources, have the capability to relocate to areas with surplus energy — whether connected or off-grid — and can address the technical, logistical and political challenges necessary to realize “moonshot” aspirations. And while the industry has long recognized that not all electricity consumers have equal footing at the bargaining table, the emergence of these powerful buyers necessitates reconsideration of service duties and a deeper reflection on how consumer protection policies need to evolve. Yet attempts at assigning a new class of customers in traditional rate proceedings have proved challenging. State and federal coordination remains disconnected, and the elevation of AI development as a national priority risks favoring a class of consumers at the expense of the rest of us.

Regulatory hurdles to co-op cost allocation

In early 2024, Basin Electric Power Cooperative proposed a special wholesale power sales rate schedule for cryptocurrency operations and other new large loads in excess of 75 megawatts. In its Federal Energy Regulatory Commission filing, Basin Electric sought to demonstrate that certain large loads, particularly cryptocurrency operations, present a higher risk of stranded assets than more traditional large loads, such as data centers or oil and gas development, due to their more speculative and potentially transient nature. FERC rejected Basin’s data center tariff, finding that the cooperative did not provide sufficient evidence to support treating these customers differently or to substantiate the claimed difference in risk among large load types. Regulators concluded that Basin Electric did not meet its burden of proof to show that cryptocurrency loads were more likely than other similarly sized customers to be relocatable or create stranded-asset risk. As a result, FERC found the proposal unduly discriminatory and not justified under the “just and reasonable” standard and rejected it.

Ultimately, to help protect existing members from long-term rate increases and financial risks that can come with serving large, non-traditional loads like data centers, Basin Electric designed a Large Load Commercial Program, which was adopted by its board of directors in June 2025. The program ensures those new loads carry their own costs while still allowing all members to benefit from economic growth, potential cost savings and shared returns. It also helps maintain Basin Electric’s financial strength by reducing the risk of stranded assets, protecting its credit rating and guiding responsible investment, all within a clear process that supports system expansion while keeping member interests first.

More recently, Tri-State Generation and Transmission Association’s High Impact Load Tariff (HILT) was likewise denied by FERC as its coordinated effort to manage and allocate data center costs from distribution through generation ran afoul of the Federal Power Act, separating FERC’s oversight of wholesale sales from state regulation of retail rates. In its decision, FERC noted that the Supreme Court has previously ruled that retail rate regulation is not allowed, regardless of how significantly it affects wholesale rates. And, even if permitted, upstream coordination would prove challenging because most of the country’s upstream supply is mixed in a socialized regional pool, like PJM. Nevertheless, the absence of a unified regulatory framework is a factor that without a doubt affects the development of more effective guardrails, may burden communities with shouldering excess supply costs and even slow data center development.

Creating new customer classes has proven elusive. Fragmented state–federal oversight and AI’s national priority risk privileging a few consumers at everyone else’s expense.

Federal actions signal a shift in the regulatory framework

The interconnection of large retail loads has traditionally been regulated by state public utility commissions, resulting in a patchwork of rules and approval processes that vary widely across states and even among utilities within the same state. To provide better uniformity, states are building a multi‑layered regulatory framework — cost recovery, tariffs, interconnection rules, efficiency mandates, and analytical studies — to ensure large loads such as data centers pay proportionately for the infrastructure and reliability impact they create. Across the first half of 2025, DSIRE Insight found 33 states took at least one action on large‑load customer treatment, with over 85 discrete actions including tariff updates, studies and interconnection rulemakings.

This year, a novel rate structure is being floated in the Virginia general assembly that would impose a 16% marginal cost added to large loads, enabling a 3% reduction for residential rates. This proposal is not dissimilar to the historical treatment of large-load, aluminum smelters in the Pacific Northwest as noted in our Umatilla Electric Cooperative case study. The measures currently being proposed at the state level collectively aim to protect existing customers, maintain grid stability and ensure that economic development does not come at the expense of affordability. Yet, these efforts might not be moving fast enough to support the national agenda of AI prioritization. Consequently, the U.S. Department of Energy is reshaping U.S. electricity regulation — so data centers can connect faster, at larger scale and more predictably than ever before.

The DOE has taken the extraordinary step of ordering FERC to create a new, national interconnection regime for large loads — including data centers — with an April 30 deadline. This is the first time DOE has invoked this degree of authority to reshape how electricity loads — not generators — connect to the grid. In other words, this policy would bypass the state regulatory compact with utilities directly. Viewed in the most restrictive light, this initiative could disrupt the current pace of local electric distribution investment, reducing incentives for local grid modernization and potentially increasing costs and risks for remaining customers.

To address public concerns on rising costs, the White House wants some of the world’s largest technology companies to publicly commit to a new compact with broad commitments ensuring data centers do not raise household electricity prices, strain water supplies or undermine grid reliability. The yet to be disclosed compact would represent one of the most significant initiatives to influence the development of AI infrastructure without enacting direct regulatory measures that run afoul of the Federal Powers Act. This announcement follows the White House’s recent, notable request to PJM to pursue strategies for reducing electricity costs through cycle capacity auctions.

FERC's expedited rulemaking will greatly affect federal-state regulatory authority and data center development. Recent federal policies and FERC actions already signal a shift, with large loads no longer seen as incremental additions manageable through traditional state interconnection processes.

Conclusion

The candid conversations we’ve had with electric cooperatives over the past year (included in our large-load case studies) emphasize the importance of carrying out the inclusive mission of providing universal service while protecting the co-op and its membership from the emerging risks of providing that service. Without a doubt, bringing in new large-load members can boost system reinvestment, but without safeguards, poor development may leave communities facing stranded costs. The good news is that best practices for local distribution safeguards have become clearer over the past year. The remaining challenge, however, is making sure these protections carry on upstream — which is why state and federal policy developments unfolding this year will fundamentally impact the electricity marketplace for all consumers.

Although coordinated grid governance poses challenges, federal-state regulatory cooperation remains crucial. Faster, more reliable data center grid connections can be achieved through unified planning, risk sharing, flexible interconnection, transparent processes and strong coordination across all levels of government. And let’s be clear: Active policy discussions about modernizing the regulatory compact in response to data‑center growth are healthy. They’re beneficial to utilities because they create clearer, more durable rules for cost recovery, risk management and long‑term capital planning at a moment when unprecedented load growth is straining the traditional model.

First, utilities benefit because updated policies help ensure fair, transparent cost allocation, protecting utilities from being forced to socialize the massive infrastructure costs associated with data center interconnections across all ratepayers. Second, active discussions highlight the limitations but also the solution to the existing patchwork of rules that create challenges for responsibly planning for new generation, transmission, and distribution. Third, utilities gain from regulatory updates that reduce exposure to stranded asset risk. Data centers can scale quickly, relocate or cancel projects, leaving utilities holding significant capital investments. Finally, utilities welcome these discussions because they offer a rare opportunity to reassert regulatory certainty at a time when the sector must invest heavily in new infrastructure.

In short: Active policy debates will ultimately provide electric cooperatives with clearer rules, better cost protection, reduced financial risk and stronger investment signals — all of which are critical as large loads and data centers reshape the energy landscape.