The NextEra-Dominion Merger and the New Economics of AI Power

The AI factory boom has exposed a growing mismatch between the speed of data center development and the pace of utility infrastructure and regulation. Hyperscale campuses can now be financed, contracted, and announced far faster than utilities can build the generation, transmission, and grid capacity required to support them.

That is why the proposed combination of NextEra Energy and Dominion Energy is more than another large utility merger. It is an early test of whether the U.S. power sector can scale itself quickly enough to support the AI infrastructure buildout without overwhelming the regulatory systems designed to govern it.

On paper, the transaction is relatively straightforward. NextEra would acquire Dominion in an all-stock deal that the companies say would create the world’s largest regulated electric utility business by market capitalization. Dominion shareholders would receive 0.8138 NextEra shares for each Dominion share, leaving NextEra shareholders with roughly 74.5% ownership of the combined company and Dominion shareholders with about 25.5%.

But the messaging surrounding the deal reveals where both companies expect scrutiny to emerge. The merger announcement repeatedly emphasized customer affordability, reliability, infrastructure investment, and long-term economic growth, stating that the combined company would deliver an “unmatched operating platform benefiting customers.” Robert Blue, chair, president and CEO of Dominion Energy, said:

Where Utility Scale Meets AI Scale

The combined company would operate under the NextEra name, serve roughly 10 million utility customer accounts across Florida, Virginia, North Carolina, and South Carolina, own 110 GW of generation capacity, and remain more than 80% regulated. The companies also pointed to more than 130 GW of “large-load opportunities” in their pipeline: language that, in today’s market, is effectively shorthand for hyperscale cloud, AI infrastructure, advanced manufacturing, and data center expansion.

But the strategic logic behind the transaction extends far beyond the merger mechanics.

NextEra Energy is effectively buying deeper access into the most important electricity market in the global data center industry: Dominion Energy Virginia’s service territory, including Northern Virginia’s Data Center Alley.

Dominion brings entrenched hyperscale relationships, enormous projected load growth, and one of the most operationally challenging grid environments in the country. NextEra brings capital scale, generation development expertise, transmission experience, and a broad mix of gas, nuclear, renewables, and battery storage capabilities.

Together, the companies are attempting to build something larger than a traditional utility platform: an integrated energy and infrastructure system designed for the era of gigawatt-scale AI and digital infrastructure deployment.

John Ketchum, chairman, president and CEO of NextEra Energy, described the deal as:

AI Load Growth Changes the Utility Equation

The repeated emphasis on affordability and customer protection throughout the merger announcement reflects a growing reality inside the utility sector: AI-driven load growth is no longer behaving like traditional electricity demand.

Utilities historically preferred predictable, diversified expansion. Population growth, commercial development, and industrial demand could justify new infrastructure investments while spreading costs across a broad customer base.

The AI infrastructure boom changes that equation.

Hyperscale data center demand is geographically concentrated, capital intensive, politically visible, and increasingly difficult to forecast over long time horizons. AI campuses may require hundreds of megawatts of capacity, phase deployment over years, alter compute architectures midstream, delay construction, or shift workloads between regions depending on power availability, permitting timelines, economics, or interconnection constraints.

That creates a new challenge for utilities: how to rapidly build generation, transmission, substations, and grid capacity without exposing ordinary ratepayers if projected demand fails to fully materialize.

This is one of the defining issues underneath the proposed merger. The transaction is built around load growth, but the deeper issue is risk allocation.

Who pays for the substations, transmission upgrades, reserve margins, and generation capacity required to support AI campuses? Who absorbs the financial risk if a 500 MW deployment arrives late, ramps slowly, or never reaches projected utilization? And who protects residential customers from infrastructure costs tied to hyperscale expansion?

In Virginia, already the center of the global data center industry, those questions are no longer theoretical.

The Epicenter of AI Infrastructure Demand

According to the U.S. Energy Information Administration, electricity consumption in Virginia grew at an annual rate of roughly 3.1% between 2019 and 2024; more than triple the national average of 0.9%. Commercial electricity sales in the state increased by nearly 30 million MWh during that period, with data centers serving as a primary driver of the growth. Meanwhile, Data Center Map currently lists more than 600 data centers across Virginia operated by dozens of providers.

That growth has effectively turned Dominion Energy into one of the most strategically important utilities in the global AI infrastructure economy.

The company now sits at the center of enormous projected load additions while simultaneously navigating PJM transmission constraints, generation planning pressures, rate-recovery debates, and increasingly skeptical regulators and local communities.

Data center expansion can significantly increase utility rate base growth, but only if utilities can recover infrastructure investments on acceptable regulatory terms and demonstrate that costs are being allocated fairly. The challenge becomes even more politically sensitive when residential ratepayers fear they may ultimately subsidize infrastructure built primarily for hyperscale AI demand.

Yet NextEra Energy appears to view those pressures less as a warning sign than as a long-term strategic opportunity.

The company is effectively betting that Dominion’s concentration of data center load, grid complexity, and infrastructure demand becomes even more valuable when paired with a larger balance sheet, broader generation portfolio, and more vertically integrated energy platform.

Selling Scale to Regulators and Communities

NextEra Energy and Dominion Energy are clearly framing the merger around a simple argument: larger scale will allow the combined company to build infrastructure faster, finance projects more efficiently, and manage rising electricity demand at lower long-term cost.

The companies argue that a larger combined platform would improve operations, procurement, construction, financing, and grid development capabilities while enabling more cost-effective investment in generation, transmission, and reliability infrastructure.

But the announcement also reveals how carefully both companies are trying to navigate the political and regulatory pressures now surrounding AI-era infrastructure expansion.

The merger proposal included a lengthy series of commitments aimed at customers, employees, regulators, and local communities:

• $2.25 billion in bill credits for Dominion customers in Virginia, North Carolina, and South Carolina distributed over two years following the merger close
• Expanded charitable and low-income assistance programs, including an additional $10 million annually in charitable giving over five years
• Retention of dual headquarters in Juno Beach and Richmond, continued operations in Cayce, South Carolina, and employment protections for Dominion’s approximately 15,000 employees
• Increased investment capacity for grid reliability, resiliency, and storm-response infrastructure
• Expanded ability to finance and build the generation, transmission, and grid infrastructure needed to support regional economic growth and rising electricity demand
• Continued emphasis on large-load tariffs designed to ensure hyperscale and other major customers pay a larger share of infrastructure costs tied to their demand

Taken together, the commitments illustrate how politically sensitive AI-driven infrastructure expansion has become for utilities, regulators, and local communities alike.

Many of the provisions are clearly designed to address concerns around customer costs, grid reliability, employment stability, community investment, and the growing perception that utilities may be building increasingly large infrastructure programs primarily to support hyperscale data center demand.

NextEra Energy and Dominion Energy will argue that greater scale enables the combined company to build infrastructure faster, finance projects more efficiently, and improve long-term reliability.

Critics, however, are likely to ask whether the merger simply creates a larger rate base and a more powerful utility platform to finance AI and data center-driven infrastructure expansion.

For hyperscalers, colocation providers, and large-load customers, the stakes extend beyond the merger itself. Regulatory approval could ultimately include explicit customer-protection measures, cost-allocation requirements, large-load tariff structures, reporting mandates, investment benchmarks, ring-fencing provisions, or limitations on shifting infrastructure costs onto residential and small-business ratepayers.

GS-5 and the New Economics of AI Power

One of the clearest indicators of where the market is heading can already be seen in Virginia’s evolving large-load tariff structure.

In November 2025, the Virginia State Corporation Commission approved Dominion Energy’s GS-5 rate class for customers requesting 25 MW or more of capacity, including large data center deployments. The rate structure, which takes effect January 1, 2027, requires certain large-load customers to pay minimum levels of contracted transmission, distribution, and generation demand in order to help shield ordinary ratepayers from infrastructure costs associated with rapid hyperscale expansion.

The implications for the data center industry could be significant.

Developers that once treated utility reservations as strategic optionality may increasingly need stronger tenant commitments, more substantial credit backing, and more disciplined deployment schedules. Hyperscalers may face longer-duration contractual obligations tied directly to the infrastructure investments their campuses require. Colocation providers may also need to mirror utility obligations inside customer contracts, particularly as AI tenants seek extremely high-density deployments while retaining flexibility around ramp timing and utilization.

The data center industry is already familiar with demand charges, pass-throughs, power purchase agreements, and utility interconnection costs. What is emerging now is something more structurally complex: a blended infrastructure-financing model that distributes risk across utilities, hyperscalers, colocation providers, tenants, regulators, and ratepayers alike.

Data Center Contracts Become Infrastructure Finance

Future data center agreements in Dominion territory (and likely in other large-load markets influenced by the NextEra-Dominion model) are likely to place far greater emphasis on long-term financial accountability, infrastructure certainty, and risk sharing.

Several themes are already beginning to emerge:

1. Minimum demand obligations
If a customer reserves 200 MW of capacity, utilities will be increasingly reluctant to build infrastructure on the assumption that actual usage may remain substantially below that level. Dominion’s GS-5 minimum billing structure offers an early template for how utilities may protect themselves against stranded infrastructure risk.

2. Longer contract durations
The substations, transmission systems, and generation assets required to support hyperscale campuses are designed to operate for decades. Regulators are unlikely to support large infrastructure investments backed by short-term or weakly committed data center loads. Service agreements aligned with 10-to-15-year utility planning horizons may become increasingly common.

3. Stronger credit requirements
Hyperscalers with large balance sheets and investment-grade credit profiles may gain a significant advantage. Smaller developers and speculative project entities could face growing requirements for parent guarantees, letters of credit, deposits, or other financial assurances.

4. Structured ramp schedules and penalties
Utilities increasingly need certainty around when load will actually materialize, not simply how much capacity may eventually be required. Contracts may evolve toward phased deployment milestones, financial penalties tied to delays, and stricter treatment of speculative queue reservations.

5. More explicit infrastructure cost allocation
Regulators will increasingly demand clarity around which grid upgrades directly support specific campuses and which qualify as broader public-benefit infrastructure investments. That distinction will heavily influence future cost recovery decisions.

6. Operational flexibility requirements
Utilities and grid operators are showing growing interest in whether large AI campuses can curtail noncritical workloads, shift compute timing, integrate onsite storage, or participate in demand-response programs during periods of grid stress.

The broader implication is that data center power agreements are beginning to evolve beyond traditional real estate-style leases with utility pass-throughs. Increasingly, they resemble long-duration industrial infrastructure contracts built around utility-grade credit, capacity guarantees, regulatory oversight, and shared infrastructure risk.

The Utility as Infrastructure Counterparty

Dominion Energy already controlled one of the most valuable assets in the AI infrastructure economy: concentrated hyperscale load growth.

What NextEra Energy brings is a far broader infrastructure platform.

In addition to owning Florida Power & Light, NextEra also controls NextEra Energy Resources, which the company describes as the largest energy infrastructure development business in the United States, spanning natural gas, nuclear, renewables, battery storage, and transmission development.

As Data Center Frontier’s ongoing coverage has repeatedly shown, the ability to combine utility service, generation development, transmission planning, storage, and long-term energy procurement into a more unified offering is becoming an increasingly important competitive advantage in the AI infrastructure market.

A combined NextEra-Dominion platform would be better positioned to offer integrated energy solutions that could include regulated utility service, behind-the-meter generation, renewable and storage PPAs, gas-backed or nuclear-supported capacity, and coordinated transmission development.

That does not mean every hyperscale customer receives a fully bespoke energy campus. It does mean the combined company could increasingly negotiate from a position of broader infrastructure control and potentially solve larger portions of the customer’s power and deployment challenges internally.

It also signals a shift in negotiating dynamics. As utility infrastructure becomes more constrained and capital intensive, hyperscalers and developers may encounter fewer contractual flexibilities and greater demands for long-term certainty, financial commitments, and load predictability.

AI Infrastructure Enters a New Power Era

Hyperscalers may ultimately be the best positioned to adapt to this new environment.

Companies such as Microsoft, Google, Amazon, Meta, and Oracle possess the balance sheets, procurement sophistication, and long-term infrastructure planning capabilities needed to operate within more rigid utility frameworks.

In some cases, hyperscalers may even prefer more structured utility arrangements if they reduce interconnection uncertainty and improve deployment visibility. But those companies will also face growing pressure to align utility commitments with rapidly evolving AI hardware cycles, chip roadmaps, deployment schedules, and internal carbon targets.

The timelines governing AI infrastructure development increasingly move at a different pace than the timelines required for utility planning, transmission development, generation construction, and regulatory approval.

Smaller developers and more speculative projects may face a much more difficult environment.

As utilities place greater emphasis on financial certainty and infrastructure accountability, land acquisition and zoning approvals alone will no longer be sufficient. Utilities and regulators will increasingly demand evidence of committed customers, credible load ramps, long-term power requirements, and meaningful financial backing before reserving scarce infrastructure capacity.

Data Center Frontier’s reporting over the past several years increasingly points toward the same conclusion: power for AI infrastructure must now be simultaneously available, affordable, reliable, scalable, clean enough to satisfy corporate sustainability targets, firm enough for AI workloads, and politically acceptable to regulators and ratepayers.

The proposed NextEra Energy-Dominion Energy combination can ultimately be viewed as an attempt to build an energy platform large enough to operate within those constraints.

Increasingly, the central question facing AI infrastructure development is no longer simply: “Can you deliver the power?” It is becoming: “What level of certainty, and what price, are customers willing to commit to in order to secure it?”

At Data Center Frontier, we talk the industry talk and walk the industry walk. In that spirit, DCF Staff members may occasionally use AI tools to assist with content. Elements of this article were created with help from OpenAI's GPT5.

Keep pace with the fast-moving world of data centers and cloud computing by connecting with Data Center Frontier on LinkedIn, following us on X/Twitter and Facebook, as well as on BlueSky, and signing up for our weekly newsletters using the form below.