Tech Giants Pour Billions Into Nuclear Restarts as Critics Question the AI Energy Rush

Is Big Tech's rush into nuclear power getting ahead of the actual technology and grid realities?

Meta, Microsoft, Amazon and Google have collectively committed tens of billions of dollars to nuclear power projects meant to feed the electricity needs of AI data centers. In January 2026, Meta announced agreements with Vistra, TerraPower, Oklo and Constellation Energy to secure up to 6.6 gigawatts of nuclear capacity, including funding for two TerraPower reactor units and a 1.2 gigawatt nuclear campus with Oklo in Pike County, Ohio, enough combined capacity to power roughly 5 million homes. Separately, Constellation Energy is expected to restart the Three Mile Island plant to supply Microsoft's data centers starting in 2027.

The push is driven by numbers that have alarmed grid planners: AI data centers consume three to five times more electricity per square foot than traditional data centers, and global data center electricity demand is projected to double by 2027, with AI workloads responsible for more than 60 percent of that growth. Wholesale power prices near hyperscale facilities have already surged 267 percent since 2020. Nuclear offers round-the-clock, carbon-free power that solar and wind can't match on their own, which is why tech companies are pursuing 20-year agreements to restart shuttered plants and fund small modular reactor startups rather than rely solely on public grids.

But most of these projects are years from delivering power. Small modular reactor designs remain largely unproven at commercial scale, and reactor restarts and new builds take years to permit and construct, timelines that don't match the immediacy of today's grid strain. Some environmental groups have also raised concerns that billions redirected toward nuclear could otherwise go to faster-to-deploy solar, wind and storage projects.

Arguments for:

  • Nuclear provides steady, 24/7 carbon-free power that intermittent renewables can't match, directly addressing AI's round-the-clock demand.

  • Locking in dedicated nuclear capacity through long-term agreements keeps AI's added electricity demand off the shared public grid, which could shield ordinary ratepayers from some cost increases.

  • Restarting shuttered plants and funding small modular reactors could revive a domestic nuclear industry that had stalled for decades, potentially lowering costs over time.

Arguments against:

  • Reactor restarts and new builds take years, doing little to address today's grid strain, the kind of pressure that already pushed New York to pause new AI data center construction.

  • Diverting billions toward nuclear could crowd out solar, wind and storage projects that can be deployed faster.

  • Wholesale power prices near hyperscale facilities have already jumped 267 percent since 2020, and critics worry households sharing the same regional grids will still absorb rising costs even as tech firms claim to bypass public utilities.

What's your take?

Is Big Tech's rush into nuclear power getting ahead of the actual technology and grid realities? Yes ↑ No ↓ Other ◇

Sources:

#NuclearEnergy #ArtificialIntelligence #DataCenters #EnergyPolicy #BigTech

Now let's hear from you.

  • You get one Take and 3 ratings so use them well.

  • Strong arguments beat loud ones.

  • Don't forget to rate your own Take.

  • Please don't feed the trolls.

About Square One

Why Microsoft And Amazon Are Turning To Nuclear Power For AI
AI’s soaring energy demand is pushing Microsoft and Amazon toward nuclear power, reshaping data center strategy, electricity markets, and the future of baseload energy.
Forbes