AI is usually discussed through chips, models and cloud platforms. But the next bottleneck may be more physical: electricity.

That is what makes NextEra Energy’s proposed acquisition of Dominion Energy strategically important. On paper, it is a utility merger. In reality, it is a bet that the AI economy will need not only more computing capacity, but also more generation, transmission, grid reliability and regulated infrastructure.

In May 2026, NextEra announced an all stock deal to acquire Dominion Energy for approximately US$66.8 billion. If completed, the combined company would serve around 10 million utility customer accounts, own about 110 GW of generation, and be more than 80% regulated. NextEra and Dominion also said the transaction would create the world’s largest regulated electric utility business by market capitalisation. The deal still requires shareholder and regulatory approvals.

The timing matters. Data centres are becoming one of the fastest growing sources of electricity demand. The International Energy Agency estimates that global data centre electricity consumption could roughly double to around 945 TWh by 2030, growing at about 15% per year from 2024 to 2030. That is more than four times faster than growth in electricity consumption from all other sectors.

This changes the strategic value of utilities. In the last decade, the AI infrastructure conversation was dominated by semiconductors, cloud capacity and model development. In the next decade, the constraint may increasingly shift to physical infrastructure: grid connections, substations, transmission lines, renewable capacity, gas generation, storage and permitting.

Dominion is especially relevant because of its exposure to Virginia, including Northern Virginia’s data centre corridor. Reuters described the proposed merger as a bet on AI driven data centre demand, noting that the deal would position the combined company to serve rapidly expanding power needs in high demand regions such as “Data Center Alley.”

That is the core M&A logic. NextEra is not simply buying another utility customer base. It is buying exposure to one of the most important electricity demand growth pools in the US.

The deal thesis has three layers.

First, regulated scale. Utilities are capital intensive businesses. Larger platforms can often finance infrastructure programmes more efficiently, spread operating capabilities across a bigger base, and support larger grid and generation investments. NextEra and Dominion have positioned the combination as a platform for scale in operations, procurement, construction and financing.

Second, load growth. Data centres create unusually concentrated electricity demand. For utilities, this can be attractive if new load is matched with appropriate tariffs, long term contracts and infrastructure cost recovery. It can also be risky if residential customers end up absorbing too much of the cost of grid expansion. That is why the regulatory design around data centre load will matter as much as the demand itself.

Third, infrastructure capability. Supplying AI data centres is not just about producing more power. It is about building the infrastructure needed to deliver that power reliably. NextEra brings large scale generation and development experience. Dominion brings a strategic regional footprint. Together, the companies are trying to position themselves as a larger energy infrastructure platform for a market where electricity demand is rising again after years of slower growth.

The risk is that this is not a simple shareholder story. Utility M&A is always political because customers, regulators and states care about affordability, reliability and control. Reuters reported that the deal may hinge on concerns over consumer electricity bills, especially as rising demand has already put pressure on power costs in some markets. To address affordability concerns, the companies have proposed US$2.25 billion in customer bill credits over two years after closing.

That creates the central tension of the transaction. AI data centres need power, and utilities need capital to build that power. But regulators will ask who benefits, who pays, and whether the merger improves outcomes for ordinary customers.

This is what makes the deal a useful signal for M&A more broadly. The next wave of AI linked M&A may not only involve software companies, chip designers or cloud platforms. It may involve utilities, grid operators, power developers, storage providers, gas infrastructure and transmission assets.

The strategic rationale is simple: if AI demand is real, electricity becomes more valuable. And if electricity becomes more valuable, the companies that control regulated power infrastructure become more strategic.

NextEra’s proposed acquisition of Dominion captures that shift. It shows that the AI economy is moving from digital ambition to physical constraint. The winners will not only be the companies with the best models or the fastest chips. They may also be the companies that can secure, build and deliver reliable power at scale.

‍