IMPLICATOR
John Ketchum chose the word "scale" in a May 18 merger release issued from Juno Beach and Richmond. The release listed about 10 million utility customer accounts, 110 gigawatts of generation, more than 130 gigawatts of large-load opportunities and $2.25 billion in proposed customer credits. The Financial Times used the $420 billion figure for the combination. S&P Global split the accounting: $66.8 billion for the stock acquisition, about $420 billion for enterprise value.
Across these deals, the premium is showing up first on capacity: power plants, cooling systems, data-center space and scarce AI talent. Capacity sets the price. Ketchum made the rationale plain: "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. It enables us to buy, build, finance and operate more efficiently, which translates into more affordable electricity for our customers in the long run."
Key Takeaways
- AI dealmaking is shifting toward capacity: power, cooling, data-center space and scarce AI staff.
- NextEra and Dominion put 10 million customer accounts, 110 GW of generation and a 130-plus GW load pipeline into the deal math.
- Cooling and data-center assets now carry premiums, from CoolIT's $4.75B sale to Aligned's roughly $40B enterprise value.
- License-and-hire talent deals still matter, but they sit beside much larger bets on infrastructure.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
The grid moved to the term sheet
NextEra and Dominion said the combined company would be more than 80 percent regulated and would operate across Florida, Virginia, North Carolina and South Carolina. It would also carry a rate base of $138 billion, with regulatory capital employed expected to grow around 11 percent annually through 2032. Robert Blue, Dominion's chief executive, described the deal as giving the company the "scale and balance sheet to deliver the generation, transmission and grid investments our customers and economies need."
NextEra proposed $2.25 billion in bill credits for Dominion customers over two years after close. The same release described more than 130 gigawatts of large-load opportunities and a 9 percent-plus adjusted earnings-per-share growth target through 2032.
Matt McClure, Goldman Sachs' global co-head of investment banking, told the FT that "AI has become a tailwind for dealmaking, and equity markets more broadly." He also said the demand was creating "a cascading impact across more traditional industries," with power companies and data-center equipment suppliers among the biggest beneficiaries.
The grid comes first.
Cooling became a premium asset
Ecolab's CoolIT deal shows how far the repricing has moved beyond utilities. Ecolab agreed to pay about $4.75 billion in cash for CoolIT, whose valuation had been $270 million three years earlier, according to the FT. The price represented 29 times estimated next-12-month adjusted EBITDA and 24 times estimated 2027 adjusted EBITDA.
CoolIT designs coolant distribution units, cold plates and direct-to-chip cooling systems, and it runs Liquid Lab R&D centers in Calgary and Taipei. Beck said: "By bringing together CoolIT's engineered cooling technologies with Ecolab's expertise in water, chemistry and digital service, we can provide our customers a complete cooling solution that improves performance and reliability while reducing water and energy use."
"The infrastructure required to run and operate an AI data centre just requires more power at the chip," Heath Monesmith, Eaton's electrical-sector president and chief operating officer, told the FT. "The whole industry needs to move at the speed of chip design."
Follow the AI capacity trade
Strategic AI news from San Francisco. No hype, no "AI will change everything" throat clearing. Just what moved, who won, and why it matters. Daily at 6am PST.
No spam. Unsubscribe anytime.
Aligned Data Centers drew an enterprise value of about $40 billion, while AIP set a $30 billion equity target that could reach $100 billion including debt.
Talent deals use a different workaround
Google agreed to pay $2.4 billion in license fees for nonexclusive rights to some Windsurf technology while hiring Varun Mohan, Douglas Chen and some R&D staff into DeepMind, without taking control of Windsurf. Microsoft paid about $650 million to license Inflection's models and hire Mustafa Suleyman, Karen Simonyan and most of the startup's staff. Reuters reported that Meta would take a 49 percent Scale AI stake for $14.3 billion, valuing the company at $29 billion, and bring Alexandr Wang into its superintelligence work.
Meta said: "We will deepen the work we do together producing data for AI models and Alexandr Wang will join Meta to work on our superintelligence efforts." Google described the Windsurf hires as adding "some top AI coding talent" to DeepMind. Jim Ryan, a Morrison Foerster partner, told the FT that these structures "can offer valuable opportunities for founders and employees," but "can also raise questions for investors."
Private capital is buying the bottleneck
S&P Global counted a record $250 billion flowing into private infrastructure funds last year, while Apollo estimates nearly $3 trillion will be required for AI infrastructure through 2028. Marc Rowan, Apollo's chief executive, told investors: "The demand for capital from this global industrial renaissance that we're going through is just off the charts."
Blackstone has a $5 billion plan with Google to build a neocloud from the ground up, using Google's proprietary processors. Jas Khaira, the Blackstone executive leading the partnership, told the FT: "Winners haven't been set yet. We are building at cost. This is the biggest cycle in capital in my entire career." Larry Fink put the constraint at CERAWeek another way: "we're going to run out of electricians as we build out data centers."
The FT noted resistance from communities blaming data centers for higher electricity bills, threatened jobs and disruption. The permit queue, too. Trump has already told tech companies to build their own power plants rather than push costs onto households. If that backlash keeps hardening, the next merger filings are likely to face pressure for numbers, not slogans.
Ketchum kept returning to scale because this corner of the market is rewarding it. NextEra and Dominion's joint proxy statement is the next public document that has to turn that scale argument into regulatory numbers.
Frequently Asked Questions
Why is AI changing merger activity?
AI systems need power, cooling, data-center space, chips, and specialist staff at a scale classic software companies did not. That pushes buyers toward utilities, data-center operators, cooling suppliers, and license-and-hire talent deals, not only model startups.
What does the NextEra-Dominion deal show?
The deal shows how electricity demand has become part of AI strategy. The proposed company would serve about 10 million utility customer accounts, own 110 GW of generation and pursue more than 130 GW of large-load opportunities.
Why does liquid cooling matter for AI M&A?
Dense AI racks produce heat that air cooling often cannot handle efficiently. That makes companies such as CoolIT and Boyd Thermal more valuable because their cold plates, coolant systems, and direct-to-chip designs sit between AI demand and physical limits.
Are acquihires replacing traditional acquisitions?
Not entirely. They are a workaround for talent and technology, as with Google-Windsurf and Microsoft-Inflection. But they do not solve power, cooling, or data-center capacity, which is why infrastructure deals are getting larger.
What is the risk for these AI infrastructure deals?
Ratepayer backlash, permitting delays, equipment shortages, and uncertain AI revenue can all pressure the deals. If communities resist data centers or regulators scrutinize power costs, buyers may have to prove savings and capacity benefits in filings.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
Marcus Schuler
Editor-in-Chief and founder of Implicator.ai. Former ARD correspondent and senior broadcast journalist with 10+ years covering tech. Writes daily briefings on policy and market developments. Based in San Francisco. E-mail: [email protected]
