NextEra’s $67B Dominion Deal Exposes Energy as AI’s New Bottleneck
The first utility to publicly prioritize AI compute loads in long-term capex planning reveals grid capacity—not chips or talent—may now determine which AI companies scale.
NextEra Energy’s $67 billion all-stock acquisition of Dominion Energy, announced Monday, marks the first time a major U.S. utility has explicitly restructured itself around AI infrastructure demand rather than traditional residential load growth. The combined entity will control more than 130 gigawatts of data center development opportunities across the Southeast, establishing power access as a previously underestimated constraint on AI deployment timelines.
The transaction, expected to close in mid-to-late 2027, combines NextEra’s renewable generation capacity with Dominion’s 51 gigawatts of already-contracted data center load tied to Amazon, Microsoft, Alphabet, Meta, Equinix, and CoreWeave, according to Fox Business. The deal creates a utility explicitly designed to serve hyperscale compute rather than adapt legacy infrastructure to accommodate it.
The Strategic Pivot
NextEra CEO John Ketchum has shifted the company’s strategy over the past 18 months from a renewable-centric model to what he calls an ‘all forms of Energy’ approach, adding natural gas and nuclear capacity to meet the 24/7 baseload requirements of AI Infrastructure, per Data Center Richness. The company plans to develop at least 15 gigawatts of new generation capacity for Data Centers through 2035.
“We are bringing NextEra Energy and Dominion Energy together because scale matters more than ever—not for the sake of size, but because scale translates into capital and operating efficiencies.”
— John Ketchum, Chairman, President and CEO of NextEra Energy
This repositioning reflects a broader inflection point in utility economics. Power demand projections in Virginia alone have tripled over the past two years, while Dominion’s data centers represented 24% of electric sales in 2023, up from 21% in 2022, according to FMI Corp. The acquired footprint spans Florida, Virginia, North Carolina, and South Carolina—states where tech companies have concentrated data center development.
The Constraint Taking Shape
Utilities received data center requests for at least 700 gigawatts of power connection development in 2025—more than the 477 gigawatts the U.S. consumed in all of 2023, per EESI. Yet Morgan Stanley forecasts U.S. data center demand will reach 74 gigawatts by 2028, with a projected shortfall of approximately 49 gigawatts in available power access.
| Year | Demand (GW) | % of U.S. Total |
|---|---|---|
| 2023 | 35 | 4.4% |
| 2028 | 74 | 6.7-12.0% |
| 2030 | — | 9-17% |
| 2035 | ~80 | — |
The Electric Power Research Institute estimates data centers could account for 9% to 17% of total U.S. electricity demand by 2030, according to Utility Dive. BloombergNEF projects total data center demand jumping from 35 gigawatts to nearly 80 gigawatts by 2035.
These projections understate the constraint’s severity. One analysis found power limitations extending data center construction timelines by 24 to 72 months, while projects relying solely on grid connections face energization delays up to 10 years in some regions due to interconnection queues, per Data Center Knowledge.
Market Signals
Pricing in wholesale power markets reflects the emerging supply constraint. PJM capacity market clearing prices for the 2026-2027 delivery year increased to $329.17 per megawatt—over ten times the $28.92 per megawatt cleared in 2024-2025, with data center growth identified as a major contributing factor. Gas turbine delivery timelines have extended to seven years, while transformer and steel supply chain constraints compound infrastructure bottlenecks.
Equipment lead times have become the practical constraint even where regulatory approval exists. Large power transformers now carry multi-year delivery schedules, while natural gas turbines—critical for baseload capacity to support always-on AI workloads—face seven-year order-to-delivery timelines. This hardware constraint means capacity announced today won’t energize until the early 2030s.
NextEra’s positioning anticipates this constraint becoming determinative. By controlling both generation assets and transmission infrastructure across high-growth Southeast markets, the combined company can offer what CEO Ketchum describes as a ‘go-to partner’ model for hyperscalers—guaranteed power access rather than speculative interconnection queue positions, according to CNBC.
Implications for AI Deployment
The strategic shift represented by this transaction establishes energy infrastructure as a tier-one constraint on AI scaling, comparable to semiconductor supply and specialized talent. Tech companies that secured power contracts early—the Amazon, Microsoft, and Meta commitments embedded in Dominion’s portfolio—gain structural advantage over competitors still navigating interconnection processes.
- Power access determines AI deployment geography, not just operating costs
- Hyperscalers with early utility partnerships gain 2-3 year time-to-market advantage
- Renewable-only strategies insufficient for 24/7 AI compute baseload requirements
- Grid modernization timelines now directly impact competitive positioning in AI
Rebecca Kujawa, head of NextEra’s renewable development unit, characterized the demand as ‘explosive,’ noting that ‘there is an enormous amount of demand being driven across the U.S. economy by the growth in data centers, driven by a lot of things of course, but specifically generative AI,’ per Bloomberg.
The constraint also creates cost pressure for traditional ratepayers. Utilities must deploy capital for grid expansion years before new capacity generates revenue, while supply chain constraints inflate equipment costs. These dynamics explain NextEra’s emphasis on scale as a mechanism for capital efficiency rather than market dominance.
What to Watch
The transaction requires regulatory approval across four states and FERC, with close expected in mid-to-late 2027. Approval timelines will signal whether regulators view utility consolidation around AI infrastructure as accelerating deployment or concentrating market power.
Competitor responses will clarify whether NextEra’s strategy represents a one-time opportunity to consolidate stranded Dominion assets or a broader template for utility sector restructuring. Southern Company, Duke Energy, and other Southeast utilities with data center exposure face strategic pressure to demonstrate comparable power access guarantees.
Interconnection queue data from regional transmission organizations will provide leading indicators of where constraints bind most severely. PJM, MISO, and SPP queue lengths and clearing prices offer quantifiable measures of supply-demand imbalance.
Finally, the composition of new generation capacity—renewable versus natural gas versus nuclear—will determine whether AI’s energy footprint aligns with corporate sustainability commitments or forces pragmatic acceptance of fossil baseload. NextEra’s shift to ‘all forms of energy’ suggests the latter outcome increasingly likely.