How AI has changed M&A
The size of deals is hitting new peaks, unloved companies are becoming sexy and PE has found a new gold mine
Up until this week, NextEra Energy, the Florida-based power company that has just unveiled swashbuckling plans for one of the largest mergers in history, was best known as America’s clean-energy champion, more focused on renewables than transformative takeovers.
Now its planned $420bn combination with its rival Dominion Energy has set out in stark relief how much the AI revolution has affected not just once sleepy utilities but the whole phenomenon of mergers and acquisitions in the US.
The boom — and what it means for sectors such as power and memory chips, as well as its broader economic impact — has transformed the scale, structure and logic of dealmaking.
It has redefined which companies matter, how transactions are structured and who finances them, as private capital investors and banks hurry to reposition themselves.
“AI has become a tailwind for dealmaking, and equity markets more broadly,” says Matt McClure, global co-head of investment banking at Goldman Sachs.
He adds that demand for the technology is creating “a cascading impact across more traditional industries”, with power companies and data centre equipment suppliers “among the biggest beneficiaries of the AI build-out”.
NextEra is a prime example. Until a few years ago, its bid for Dominion — intended to create a 240-gigawatt giant catering to data centres and other big clients — would have seemed almost unthinkable.
Utilities were viewed as too regulated, politically sensitive and vulnerable to a backlash over energy prices for large-scale consolidation.
AI, together with the Trump administration’s deregulatory drive and America First approach to antitrust, has changed that.
The AI boom is also reshaping Wall Street, the US corporate order and the real economy, as established businesses desperately seek scale to compete with trillion-dollar technology companies.
CEOs and boards have caught deal fever Fomo — fear of missing out — while their rivals transform themselves for the AI age.
Groups previously unloved by Wall Street have become the target for multibillion-dollar transactions, such as Globalstar, a satellite operator, bought by Amazon for $11.6bn last month. Others have seen their shares soar. The market capitalisation of Sandisk, the computer storage group, has increased by 4,000 per cent since going public in February last year due to insatiable demand for its chips.
Even the nature of deals is changing, with so-called hyperscalers such as Nvidia, Meta and Google spending billions to hire elite AI engineers to avoid buying whole companies and sidestep antitrust regulators. The boom is also redrawing private capital, with groups such as BlackRock, Blackstone and Apollo rushing to make AI-related investments.
To some, the momentum seems unstoppable. But two tests are looming. In the coming months, capital markets will have to absorb IPOs of a size never experienced before, with SpaceX, OpenAI and Anthropic all set to be valued at around $1tn or more apiece.
The shift to AI at the heart of the M&A upheaval is also facing backlash from Americans who blame data centres for increasing electricity bills, threatening jobs and disrupting their communities.
Scale, scale, scale
A quarter of a century ago, the internet revolution was defined by companies that could conquer the world with relatively little capital. Software groups scaled at extraordinary speed without needing factories, power plants or vast physical infrastructure beyond logistics hubs and cellphone towers.
The AI revolution is overturning that model.
Winning now requires immense scale: chips, energy, fibre networks, data centres and financing measured not in billions, but trillions of dollars.
Jensen Huang, chief executive of Nvidia, recently argued that “trillions of dollars of infrastructure still need to be built” to support the AI boom.
“The amount of computation demand for software in the past is a tiny fraction of what is necessary in the future,” Huang said on a call with investors. “And AI is here, AI is not going to go back.”
That race for scale is driving the new era of megadeals.
Utilities are pursuing transformational mergers to secure power generation, infrastructure investors are hurrying to finance data centres and technology groups are striking increasingly unconventional transactions to lock up talent and supply chains before rivals do.
The shift is being accelerated by geopolitics and Donald Trump’s return to the White House. His administration has treated AI as a strategic arms race against China, creating an environment in which corporate consolidation is viewed through the lens of national competitiveness rather than consumer protection.
On Thursday, Trump postponed signing an executive order on AI under which the sector’s leading companies would voluntarily submit their models to government checks for national security and cyber risks.
Peter Orszag, chief executive of Lazard, describes the trend as part of a broader move towards “discretionary state capitalism”, in which governments play a more active role in directing capital and industrial policy. He argues the US economy is increasingly being powered by investment tied to AI, from hyperscaler capital expenditure to construction and energy infrastructure.
The political climate has also transformed antitrust from the Biden administration’s approach. Dealmakers increasingly believe consolidation will be tolerated — or even encouraged — if it supports the administration’s broader Maga agenda of AI dominance, domestic manufacturing and economic nationalism.
The result is a corporate environment in which size itself has become a strategy.
Second life
For much of the past decade, hardware groups such as Sandisk were dismissed as low-growth relics in a market obsessed with high-margin software.
The AI boom has transformed their fortunes.
As demand for data storage and computing capacity has surged, investors have rushed into Sandisk as well as groups such as Western Digital and SK Hynix, in recognition that the large language models that dominate AI depend on vast memory and chip infrastructure.
The shift has also transformed the energy sector.
Constellation Energy, operator of the country’s largest nuclear fleet, has become a market darling as hyperscalers race to secure the power needed for data centres.
Constellation’s nearly $27bn takeover of rival Calpine is part of a wave of utility consolidation tied to AI demand, alongside BlackRock-backed deals worth a combined $39bn for power groups AES and Allete.
Industrial groups providing the physical “plumbing” of the AI era have also become strategic assets. Eaton’s $9.5bn acquisition of Boyd’s thermal division this year highlights the contests to control liquid cooling technology vital to next-generation computing facilities.
“The infrastructure required to run and operate an AI data centre just requires more power at the chip,” says Heath Monesmith, president and chief operating officer of Eaton’s electrical sector. “The whole industry needs to move at the speed of chip design.”
KKR made more than a 15-fold return on March’s $4.75bn sale of CoolIT Systems, which had jumped from a $270mn valuation just three years before. The company’s transition from a peripheral player in gaming to critical AI infrastructure illustrates how the boom is minting winners from once-niche specialists.
The ripples are felt well beyond hardware and energy. General Catalyst’s $6.3bn acquisition of travel group American Express Global Business Travel and its partnership with Nelson Peltz on an $8bn takeover of London-based asset manager Janus Henderson reflect growing bets that AI can transform service businesses through automation and productivity gains.
Hire or buy?
Big Tech has found a takeover workaround in the AI race: licensing a start-up’s intellectual property and hiring its top talent without acquiring the corporate entity.
Such “acquihires” came to prominence as a defensive manoeuvre against the antitrust policy of the Joe Biden era. They have persisted under a more deal-friendly Trump administration, helping companies sidestep the lengthy reviews that accompany traditional M&A, although state-level scrutiny has begun to intensify.
Almost every hyperscaler has adopted the playbook. Google recently deployed $2.4bn to secure Windsurf’s coding architecture and leadership, following similar raids by Microsoft on Inflection AI and Amazon on the engineering core of Adept AI.
Meta’s $14.3bn investment in Scale AI, which installed Alexandr Wang at the helm of the tech giant’s new superintelligence division, remains the benchmark for this talent-first consolidation.
Nvidia, the latest heavyweight to embrace the strategy, has carried out several such deals, including with networking start-up Enfabrica and a $20bn arrangement with its upstart chip rival Groq.
Groq sent the cash to its shareholders and shared top talent and intellectual property with Nvidia, but will continue operating its own small cloud business.
“Nvidia wanted to be really quiet about it but it was hard to keep such a large deal under wraps,” says a person with knowledge of the deal, which was announced on Christmas Eve.
Jim Ryan, a partner at Morrison Foerster, says Big Tech’s growing adoption of acquihire structures “can offer valuable opportunities for founders and employees”. But, given the complexity of such arrangements, “they can also raise questions for investors”.
AI and private capital
Just four years ago, the private capital industry appeared to be entering a bleak era as the age of easy money came to an end. Interest rate rises threatened the sector’s highly leveraged portfolios. Now, however, such asset managers appear to have embarked on a new golden age, as the chief underwriters of tens of trillions of dollars for AI.
Investors poured a record $250bn into private infrastructure funds last year as capital flooded towards data centres, power generation and digital networks, according to S&P Global.
Apollo estimates nearly $3tn will be required for AI infrastructure up to 2028, with private credit and specialist funds expected to provide much of the financing.
“The demand for capital from this global industrial renaissance that we’re going through is just off the charts,” Marc Rowan, Apollo’s chief executive, told investors late last year.
For Apollo, the immense expenditure on energy and AI infrastructure offers the opportunity to use its vast pool of life insurance assets to become a long-term lender to promising companies.
For smaller players in the private capital industry, particularly midsized buyout firms, the AI boom has been far more punishing. Valuations and margins have been compressed across large swaths of the software sector, intensifying a market shift in which scale is essential to survive.
All the while the giants are pursuing aggressive plays. Blackstone recently unveiled a $5bn plan to construct a “neocloud” from the ground up in partnership with Google.
The venture aims to deliver high-octane computing power for AI models, using Google’s proprietary microprocessors — a bet that could eventually reach a valuation in the tens of billions.
Blackstone argues that greenfield projects in emerging sectors at times can be more profitable than the traditional strategy of acquiring existing businesses at a steep premium.
“Winners haven’t been set yet. We are building at cost,” says Jas Khaira, the veteran Blackstone executive spearheading the Google partnership. “This is the biggest cycle in capital in my entire career.”
Will it end in tears?
The AI boom is good news for M&A, previously unfavoured companies, engineers hired at vast expense and the titans of private capital.
The rest of America is not so convinced. This month former Google chief executive Eric Schmidt was booed at the University of Arizona when he compared AI to the transformational impact of the computer in a commencement address.
“I know what many of you are feeling,” he added. “There is a fear in your generation that the future has already been written, that the machines are coming, that the jobs are evaporating.”
Across the American heartland, communities are mobilising against the rapid proliferation of sprawling data centres, citing the strain on local power grids as electricity bills soar. According to one recent NBC poll, AI is even less popular than the US’s disliked Immigration and Customs Enforcement agency.
Such strength of feeling may yet decide how far the AI revolution sweeps all before it — and how long the valuations premised on its success endure.
The SpaceX, OpenAI and Anthropic IPOs — all of which are due in the coming year — are designed to finance the relentless capital expenditure on the models behind the AI boom, one of the biggest peacetime investments in history. With such enormous sums at stake, some investors question the maths and the projected profits of the companies involved, evoking the memory of the dotcom bubble.
In the interim, global M&A is now dominated by the AI-inspired race to control the world’s energy, fibre networks and computing capacity.
For McClure at Goldman Sachs, the transformative power of the AI revolution is highlighted by the sheer ambition of utilities and other groups that have suddenly made their way into the limelight.
“A decade ago, these enterprises were categorised as stable, GDP-correlated performers,” he says. “The insatiable demand for data centres has pivoted them towards a trajectory of exponential expansion.”
The question is how long that exponential expansion can continue for the beneficiaries of the AI boom — and the dealmaking it has set off.