Four companies, Amazon, Meta, Google, and Microsoft, bought nearly half of all corporate clean energy contracts globally in 2025. Meanwhile, everyone else pulled back, with overall volumes declining for the first time in eight years. What looks like a sustainability win is actually a story about market concentration, economic fragility, and what happens when the clean energy transition and the U.S. economy both depend on the same handful of companies making the same massive bets. In this discussion, we break down the data and explain why CIOs, energy leaders, and policymakers should be paying close attention.

Transcript:

So here’s a very hot number that should stop you in your tracks. In 2025, four companies,  Amazon, Meta, Google, and Microsoft, accounted for nearly half of all clean energy power purchase agreements globally or ppa. Four companies. Half the market. Now on the surface, that sounds like a great headline for sustainability. Big tech is putting real money behind clean energy. But there are three things happening simultaneously that deserve a little more attention.

First, the concentration problem. When four buyers represent 49% of global clean PPA volume, you don’t have a clean energy market — you have a clean energy oligopsony. Which is the buy-side version of a monopoly. A monopoly is when one seller controls the market. An oligopsony is when a small number of buyers have so much purchasing power that they effectively control pricing, supply, and terms — even though there are plenty of sellers.
That means pricing, project selection, geographic priority — all of it gets shaped by the infrastructure needs of hyperscale data centers. If you’re a mid-sized enterprise trying to sign a clean energy PPA for your operations, you’re now competing for capacity against companies spending tens of billions on AI infrastructure. Your negotiating leverage just got a lot smaller.

Second, the AI energy story is now the clean energy story. These purchases aren’t driven by altruism. They’re driven by the fact that AI training and inference workloads are extraordinarily power-hungry, and these companies have made public carbon commitments they need to honor. And keep in mind,  these are the same four companies that are arguably the primary engine of U.S. economic growth right now, with data center buildout plans totaling hundreds of billions of dollars across dozens of states. They’re driving job creation, grid investment, construction demand, and regional economic development at a scale that rivals federal infrastructure programs. So what you’re really seeing is AI infrastructure demand becoming simultaneously the largest driver of corporate renewable energy procurement on the planet and one of the most significant forces in the U.S. economy. That’s not a sustainability initiative anymore,  that’s industrial policy being set by the private sector.

Third: and this is the part that should concern everyone — the rest of the market pulled back. Overall corporate clean PPA volumes fell 10% in 2025. That’s the first decline in eight years. So while big tech is buying more, everybody else is buying less. Why? Rising PPA prices, interest rate pressure on project economics, permitting delays, and frankly, some fatigue around long-term energy commitments in an uncertain macro environment.

And fourth: zoom out and look at the full picture. Between the data center buildout and the clean energy push, a handful of companies now account for an outsized share of U.S. economic momentum. We’re talking about capital expenditure, construction jobs, power grid expansion, land development, renewable energy deployment, all heavily concentrated in the same four names. That’s not just a tech story or an energy story. That’s a systemic risk story. If even one of these companies hits a wall,  a slowdown in AI revenue, a regulatory crackdown, a strategic pivot, a market correction, the shockwave doesn’t stay in Silicon Valley. It hits energy developers who banked on those PPA contracts. It hits the construction crews building data centers in rural Virginia and central Ohio. It hits the municipalities that rezoned land and issued permits based on projected tax revenue. And it hits the clean energy transition itself, because nearly half the global buying power just got pulled off the table.

The opportunity here is real: AI could pull forward more clean energy deployment than any policy mandate ever has. But we’ve built an extraordinary amount of economic dependence around a very small number of companies making very large bets. And history tells us that kind of concentration always looks brilliant, right up until it doesn’t.