- The White House is moving to stop hyperscalers from sticking Americans with the AI power bill
- A $15 billion emergency baseload strategy could reshape grid economics – yay!
- A simultaneous rollback of U.S. greenhouse gas authority is bad for everyone – boo!
2026 has been a good-news, bad-news year for energy policy in the United States.
On the one hand, the White House is moving to stop hyperscalers from sticking Americans with the AI power bill, backing a $15 billion emergency power strategy that could reshape grid economics. On the other hand, it has rolled back the core U.S. federal power to restrict greenhouse gas emissions. That combination — fiscal toughness on grid costs and deregulation on emissions — creates one of the sharpest policy contrasts of the AI era.
Let’s start with the glass-half-full news: the Trump administration is currently exploring a new energy cost-sharing compact with hyperscale cloud and AI operators that would prevent utilities and hyperscalers from sticking customers with the massive bill for the total overhaul of transmission, generation, and reliability required to power America’s vast AI data center buildout.
“I never want Americans to pay higher electricity bills because of data centers,” President Trump said on January 12 on Truth Social.
A few days later, the U.S. Department of Energy outlined a plan for Emergency Capacity Auctions that would require tech giants to sign 15-year “take-or-pay” contracts, covering the cost of new power plants regardless of usage. This $15 billion plan could add 7.5 gigawatts of baseload power while shielding residential ratepayers from AI-driven energy costs, according to a research note from Jeffries.
As government plans go, this is a “check out the big brain on Brad” level strategy — in other words, I like it.
The drawback? The deal is voluntary. So, will the hyperscalers get in line? They’re not exactly known for adhering to the honor system. We’re talking about companies including Microsoft, Amazon, Google and Meta — outfits whose AI factories are devouring gigawatts and have an entirely nebulous relationship with ethics. (And, often, reality).
In theory, the President could force the hyperscalers to play ball by issuing an Executive Order, instructing federal agencies not to issue unless companies sign on. He hasn’t. Partly because he doesn’t want a biff with the Big Tech bros, including Jeff Bezos, but also because they probably don’t have much leverage — they need the power more than the government needs the data centers.
This is potentially a big relief for the residents of Louisiana, where Entergy is attempting to socialize the full infrastructure cost of Meta’s giant new AI megacenter.
And this all somewhat reduces the risk of the type of U.S. energy apocalypse warned of by FNTV here: big electric load + greedy/cheap hyperscalers + socialized cost + giant electricity bills = unhappy citizenry and viva la revolution!
So that’s the good news.
The bad news is that, at the same time, the Trump administration has rescinded the EPA’s 2009 greenhouse gas “endangerment finding” — the legal foundation for federal climate controls — effectively weakening federal authority to regulate emissions just as power demand explodes. That rollback dismantles the scientific and legal basis for regulating greenhouse gases under the Clean Air Act, and it appears to be primarily aimed at currying favor with flat-Earthers and people who don’t think Neil Armstrong went to the moon.
If you’ve been following my reporting on the world’s new network stack, you know AI’s growth isn’t just about chips and software. It’s about power.
Real power. Grid power. Dollars and policy power. And the U.S. today is about 50 years and several hundred gigawatts behind where it needs to be in grid capacity.
And now we’re watching that challenge collide head-on with U.S. policy — producing what may be the defining energy paradox of the AI era.
For more of this kind of thing, follow and subscribe here.
Stephen M. Saunders MBE is a communications analyst and USPTO-registered inventor examining how digital infrastructure — 5G, cloud, and AI —is reshaping industry, power and society, as well as underpinning the emerging, ubiquitous global digital economy. As anchor of FNTV and a longtime industry insider, he focuses less on growth narratives and more on execution, risk and how hyperscale technology is distorting markets, governance and society at scale.
Opinion pieces from industry experts, analysts or our editorial staff do not necessarily represent the opinions of Fierce Network.