Meta, Amazon and Google back next-gen nuclear projects to power AI data centers

Meta, Amazon, and Google are committing billions to fund small modular nuclear reactors to power their AI data centers. Their long-term purchase agreements give nuclear developers the revenue certainty banks require before financing construction.

Categorized in: AI News Finance
Published on: Apr 12, 2026
Meta, Amazon and Google back next-gen nuclear projects to power AI data centers

Big Tech's Nuclear Bet: Funding Advanced Reactors to Power AI Data Centers

Meta, Amazon, Google, and other technology companies are committing billions to develop small modular nuclear reactors, betting that next-generation power plants can meet surging electricity demand from AI infrastructure. The deals mark a shift in nuclear financing, moving the sector beyond its traditional reliance on government support and venture capital.

Meta agreed in January to fund two TerraPower units capable of producing 690 megawatts and signed a separate deal with Oklo for a 1.2 gigawatt nuclear campus in Ohio. Amazon is partnering with X-energy to deploy more than 5 gigawatts of small modular reactors by 2039, while Google signed an agreement with Kairos Power to bring online its first reactor by 2030.

Why Tech Companies Are Moving Into Nuclear

U.S. electricity consumption is expected to grow 1% this year and 3% next year, driven primarily by data center expansion, according to the Energy Information Administration. Data centers require constant, reliable power-a constraint that makes nuclear attractive despite the technology's historical financing challenges.

Tech companies' involvement solves a critical problem for nuclear developers: revenue certainty. Banks require long-term power purchase agreements before financing construction. By committing to buy electricity for decades, Meta and Amazon provide the revenue visibility that traditional lenders demand, said Shioly Dong, senior analyst at BMI, a unit of Fitch Solutions.

Why Small Modular Reactors Appeal to Investors

Small modular reactors are smaller, more advanced, and faster to build than conventional nuclear plants. Their modular design and shorter construction timelines reduce upfront capital exposure-a key advantage for institutional investors wary of the nuclear sector's history of cost overruns and delays.

Tim Winter, portfolio manager of the Gabelli Utilities Fund, said the sector needs "someone" to absorb the risks of construction delays and cost inflation. "The degree the hyperscalers are willing to do that will determine just how much of a boost these agreements give the sector," he said.

Banking Interest Emerging, But Cautiously

Long-term power purchase agreements are attracting institutional investor attention to a sector historically dependent on government subsidies. Tess Carter, associate director of the energy and climate practice at Rhodium Group, said banks are beginning to express interest in nuclear financing-a development that would represent a significant shift.

However, institutional investors remain cautious. High construction and technology risks mean capital is not flowing into the sector at scale yet. None of the companies developing these reactors have begun commercial electricity production.

The Execution Challenge Ahead

Advanced nuclear still faces significant hurdles. Licensing, fuel supply, construction, and financing all require flawless execution-and demand for power alone won't solve these problems, said Bonita Chester, an Oklo spokesperson.

A looming skills shortage presents another constraint. Competition from data centers and other industries for electricians, pipefitters, and other skilled workers could bottleneck deployment as companies attempt to scale production, according to a recent report by the Nuclear Scaling Initiative.

Tech companies' financial backing has created momentum. Whether it translates to actual power generation depends on whether the industry can execute on multiple fronts simultaneously.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)