$67 billion! AI’s rise fuels the largest energy merger and acquisition deal in U.S. history
TechFlow Selected TechFlow Selected
$67 billion! AI’s rise fuels the largest energy merger and acquisition deal in U.S. history
GPUs burn computational power, the power grid burns money, and ultimately, ordinary consumers foot the bill.
Navigating Web3 tides with focused insightsAuthor｜Hualin Dance King
Editor｜Jingyu
A few years ago, if someone told me that AI would ultimately reshape America’s electricity landscape, you probably wouldn’t have taken it seriously. After all, we’re talking about software, algorithms, and model parameters—things that sound utterly disconnected from power plants.
But on May 18, 2026, that perception was shattered entirely.
NextEra Energy announced its $67 billion acquisition of Dominion Energy—the largest utility merger in U.S. history.
The figure itself is staggering. Yet what matters more is the underlying logic: this deal wasn’t driven by traditional energy strategy, but by the insatiable, unquenchable thirst for electricity from global AI data centers.
01 The Invisible “Computing Power Arteries”
To understand this merger, you need to know one place—Northern Virginia’s Loudoun County, widely known in the industry as “Data Center Alley.”
This seemingly unremarkable region hosts the world’s densest concentration of data centers. AWS, Microsoft, Google, and Meta house vast numbers of servers here. It’s estimated that roughly 70% of global internet traffic passes through this area daily. And Dominion Energy is the primary electricity provider for this region.
Dominion holds over 51 GW in signed data center power contracts—a figure equivalent to the generating capacity of approximately 50 large nuclear power plants—and that number continues to grow. Load demand in Dominion’s Virginia service territory is projected to surge by 121% by 2045.
That’s why NextEra was willing to spend $67 billion—not to acquire a conventional utility company, but to secure the scarcest resource of the AI era: “power supply rights” located physically close to computing cores.
Markets spent two years pricing AI chips; now they’re beginning to price the grid.
02 A Grid Under Strain
Viewed within the context of the past year, this merger isn’t an isolated event—it’s the latest link in a chain reaction.
Rewind to 2025: the International Energy Agency (IEA) had already sounded the alarm.
In 2025, global data center electricity demand surged by 17%, while overall global electricity demand grew only 3%. Demand growth from AI-dedicated data centers far outpaced the broader market—running solo at full throttle. The IEA forecasts that by 2030, global data center electricity consumption will double from 415 TWh in 2024 to roughly 945 TWh. Most of that additional 530 TWh will be attributable to AI training and inference workloads.
The five major tech giants collectively spent over $400 billion in capital expenditures in 2025, with a substantial portion flowing into data center construction—and that figure is expected to rise another 75% in 2026.
The grid is buckling under pressure.
Just two days before this merger was announced—on May 16—a report from Monitoring Analytics revealed an unsettling reality: electricity prices in PJM Interconnection, the largest U.S. power market, surged by a staggering 76%—a rise described as “irreversible.” PJM covers more than ten states, including Virginia, Maryland, and Pennsylvania—precisely one of the most densely concentrated AI infrastructure regions in the country.
The report’s use of the term “irreversible” was unusually stark. This isn’t describing a temporary price fluctuation subject to adjustment—it signals a fundamental shift in the electricity supply-demand structure.
Earlier still, at the end of 2025, Northern Virginia underwent a real-world stress test on its grid. Voltage fluctuations caused 60 data centers to disconnect simultaneously, instantly generating a 1,500-megawatt surplus of electricity. That sudden energy shock served as a stark reminder of how fragile AI infrastructure is to grid stability—and how demanding it is on power supply.
03 NextEra’s Bet
NextEra is no ordinary traditional utility. It is the largest wind and solar power generator in the United States, with deep expertise in building and operating clean-energy infrastructure. Acquiring Dominion goes well beyond simple scale expansion.
The true strategic value lies in combining NextEra’s clean-energy generation and energy storage capabilities with Dominion’s dominant market position in the Data Center Alley corridor.
Jigar Shah, former head of the Department of Energy’s Loan Programs Office, put it plainly: applying NextEra’s energy storage expertise to data center load in Virginia “could be transformative”—because data centers don’t just need electricity; they need stable, predictable electricity—ideally electricity that can be stored during off-peak hours and deployed on demand.
What NextEra is betting on is that AI compute demand won’t slow down.
Given current investment trends, this bet isn’t reckless. Through mechanisms like “large-load tariffs,” major electricity consumers—i.e., data centers—will directly participate in financing infrastructure development. This means NextEra’s future capital burden for expanding transmission lines and power generation facilities can be partially shifted onto tech companies—rather than borne entirely by the utility itself.
Of course, regulatory hurdles loom large.
Acquiring Dominion would make NextEra a multi-state electricity super-giant, likely triggering rigorous scrutiny from state public utility commissions across its footprint. Consumer advocacy group Clean Virginia has already issued a public warning, calling for the “strictest possible review” of the deal, citing concerns over consolidation of control over Virginia’s electric grid.
04 Who Pays the Bill?
When AI consumes electricity at breakneck speed and electricity prices skyrocket—who ultimately foots the bill? That may be the most critical question behind this landmark acquisition.
Building electricity infrastructure costs money—and that money ultimately flows into electricity rates. Some U.S. utilities have already begun implementing “construction-work-in-progress” (CWIP) financing mechanisms, allowing them to charge consumers before projects are even completed. In other words, residential customers begin paying for data center infrastructure—even before they receive any benefit from new electricity capacity.
PowerLines’ analysis delivers a sobering figure: residential consumers could shoulder approximately $700 billion in costs associated with AI-driven electricity infrastructure investments, gradually passed on via higher electricity bills.
$700 billion. That’s on par with tech companies’ capital expenditure scale—but flows in the opposite direction. Tech firms’ $400 billion in capex yields shareholder returns, improved model capabilities, and competitive advantage. Meanwhile, consumers’ share of these costs translates solely into ever-rising electricity bills.
There’s a structural inequity embedded in this merger’s logic—and indeed, throughout the entire AI infrastructure investment wave.
Data centers are private assets. Economic gains from AI accrue almost exclusively to tech companies and their shareholders. Yet the grid enabling all of this—the public infrastructure—is built and maintained at shared cost borne by every user. This isn’t a new issue—but AI has magnified it to an unprecedented scale.
A $67 billion acquisition has laid bare, for the first time, the energy industry’s consolidation logic: AI’s prosperity doesn’t unfold solely inside data centers. It spreads outward along cables—into the grid, onto utility balance sheets, and finally onto every ordinary household’s electricity bill.
This merger is not the end. Given the current pace of AI compute expansion, it’s likely only the beginning—the restructuring of the electricity landscape has just begun.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
极客公园
