Author | Hu Lin Dance King

Editor | Jing Yu

If someone had told me a few years ago that AI would ultimately reshape the electricity landscape in the United States, I probably wouldn't have taken it seriously. After all, we are talking about software, algorithms, model parameters—these things sound unrelated to power plants.

But on May 18, 2026, this understanding was completely shattered.

NextEra Energy announced its acquisition of Dominion Energy for $67 billion, setting a record for the largest utility merger in U.S. history.

This figure is shocking enough, but what is even more concerning is the underlying logic—what drove this transaction was not some traditional energy strategy, but the insatiable thirst for power from global AI data centers.

01 The Invisible "Computing Power Veins"

To understand this merger, you first need to know a place—Northern Virginia, particularly around Loudoun County, known in the industry as "Data Center Alley."

This area is home to the most densely populated cluster of data centers in the world, with a significant number of servers from AWS, Microsoft, Google, and Meta hidden in this seemingly ordinary land. It is estimated that about 70% of the world's internet traffic passes through here every day. And Dominion Energy is the main power supplier for this region.

Dominion has more than 51 GW of data center contract demand—what does 51 GW mean? It is roughly equivalent to the installed capacity of 50 large nuclear power plants, and this number is still growing. The load in Virginia's DOM area is expected to increase by 121% by 2045.

This is the core reason why NextEra is willing to fork out $67 billion—not to acquire a traditional power company but to buy the most scarce resource in the AI era—the "power supply rights" close to the core of computing power.

The market has spent two years pricing AI chips; now it is starting to price the power grid.

02 The Struggling Power Grid

If you place this merger on the timeline of the past year, you will find it is not an isolated event, but the latest link in a chain reaction.

Going back to 2025, data from the IEA sounded the alarm.

In 2025, global data center electricity demand surged by 17%, while the overall growth in global electricity demand was only 3%. The growth rate of AI-specific data centers far outpaced the market. The IEA predicts that by 2030, global data center electricity consumption will double from 415 TWh in 2024 to about 945 TWh—most of the added 530 TWh is attributed to AI training and inference workloads.

The total capital expenditures of the five major tech giants in 2025 exceeded $400 billion, with a significant portion flowing toward data center construction, and this number is expected to grow by another 75% in 2026.

The power grid is beginning to struggle.

Just two days before this merger was announced, on May 16, a report from Monitoring Analytics revealed a disturbing reality: prices in the United States' largest electricity market, PJM Interconnection, experienced a "irreversible" significant increase, with a rise of up to 76%. The PJM area covers more than ten states, including Virginia, Maryland, and Pennsylvania, which happens to be one of the most densely populated regions for AI infrastructure.

The report uniquely used the term "irreversible." This is not indicating an adjustable price fluctuation; it is stating that the power supply and demand structure has undergone a fundamental change.

Earlier, at the end of 2025, a real stress test had already occurred in Northern Virginia's power grid. Voltage fluctuations caused 60 data centers to disconnect simultaneously, instantly creating a power surplus of 1500 megawatts—this momentary energy shock reminded everyone how fragile the AI infrastructure is against the stability of the power grid and how demanding its power supply requirements are.

03 NextEra's Bet

NextEra is not an ordinary traditional power company. It is the largest wind and solar power producer in the United States, with deep experience in the construction and operation of new energy infrastructure. This acquisition of Dominion is not just simple scale expansion.

The true strategic value of this deal lies in combining NextEra's clean energy and storage capabilities with Dominion's market position in the data center corridor.

Former loan project director at the Department of Energy, Jigar Shah, put it plainly—he believes applying NextEra's storage expertise to Virginia's data center loads "could be transformative”—because data centers not only need electricity, but also stable, predictable power, ideally electricity that can be stored during low demand periods.

NextEra is betting that the demand for AI computing power will not stop.

Based on current investment trends, this bet is not aggressive. Through the "large load tariff" mechanism, large electricity users (i.e., data centers) will directly participate in infrastructure financing, which means the capital pressure for NextEra to expand transmission lines and power generation facilities can partially be passed on to tech companies—rather than being entirely borne by utility companies.

Of course, regulatory challenges are also ahead.

Acquiring Dominion means NextEra will become a superpower in electricity across multiple states, potentially facing rigorous scrutiny from state utility commissions. Consumer rights organization Clean Virginia has already publicly issued a warning, calling for the "strictest review" of this transaction, raising concerns about the centralization of control over Virginia's power grid.

04 Who Pays for the Electricity Bill?

When electricity resources are crazily consumed by AI and electricity prices surge, who ultimately pays for the rising electricity costs? This may be the most pressing question behind this shocking acquisition.

Building power infrastructure requires money, and this money will ultimately enter electricity prices in various ways. Utility companies in some areas of the United States have started adopting a "construction financing" mechanism, allowing them to charge consumers before projects are completed. In other words, residential users start paying for the infrastructure of data centers even before they enjoy any additional electricity.

An analysis from PowerLines provided a shocking figure: AI-driven electrical infrastructure investments may cost residential consumers approximately $700 billion, gradually passed on through rising electricity bills.

$700 billion. This is on the same scale as the capital expenditures of tech companies but is directed to completely different ends. The $400 billion capital expenditure for tech companies results in shareholder returns, model capability enhancement, and corporate competitive advantages; while the costs borne by consumers only generate an increasingly high electricity bill curve.

There is a structural unfairness embedded in the logic of this acquisition and in the entire wave of AI infrastructure investment.

Data centers are private assets, and the economic benefits brought by AI are concentrated in tech companies and their shareholders. However, the power grid that supports all of this is public infrastructure, with construction and maintenance costs shared by all users. This is not a new problem, but AI has amplified it to an unprecedented scale.

The $67 billion acquisition has laid bare the logic of integration in the energy industry, clearly showing everyone that the prosperity of AI is not confined to data centers; it spreads along cables into the power grid, into the balance sheets of utility companies, and ultimately into the electricity bills of every ordinary household.

This merger is not the end. Given the current pace of AI computing power expansion, this is likely just the beginning—the reshaping of the electricity landscape has only just begun.