Data, Deals, and Government. What’s New with Nuclear?

By: Rob Cummins, Staff Writer

For most of the past two decades, America’s nuclear narrative has been defined by a frustrating paradox: the country still relies on nuclear power for large volumes of carbon-free electricity, yet new projects have struggled under cost overruns and delays that caused utilities to retreat from large, risky projects. In a June 11 article, The Wall Street Journal notes that the U.S. “produces less nuclear power than it did a decade ago,” and that repeated project failures drove a sustained backpedal by major power companies.

Over the last 18 months, however, the story has started to change. While the engineering challenges remain substantial, the customer set, financial infusions, and political incentives are being pulled forward by a unique customer that utilities haven’t faced before at scale: “city-sized” electricity demand from AI data centers.

Big Tech Becomes an Anchor Buyer

Meta’s January 2026 announcement of a nuclear-power plan to fuel its AI is emblematic: the company described itself as an “anchor customer” for both new and existing nuclear capacity to supply its artificial-intelligence data centers, backing projects with Oklo and TerraPower and committing to purchase and expand output from three Vistra nuclear plants in Ohio and Pennsylvania.

What makes this different from prior waves of corporate clean-energy procurement is the shape of what is being bought. Beyond just contracting, Meta is committing to procure long-duration, high-reliability supply, paired with investments that address major manufacturing and deployment bottlenecks, like licensing and project mobilization. Vistra’s deal includes over 2,000 MW of operating capacity and 400 MW of uprates across three sites, along with planning for federal license extensions. For context, 2000MW is enough for about 1.5 Million homes (or 50 million lava lamps, if you’re into that).

Meta is being aggressive, aiming for “first new reactors” as early as 2030 and 2032. In this historically slow-moving industry, that kind of commercial urgency is a major shift. For builders, it represents the removal of a significant commercialization barrier, since their outlook on survival hinges on promises of chunky, long-term purchasing with reliable demand.

Markets heard the signal. A WSJ markets note reported that Oklo and Vistra shares jumped on the Meta news, a visible indicator that public investors now treat data-center-linked nuclear contracts as financially material.

And nuclear’s resurgence isn’t just a greenfield build story: the fastest capacity gains are increasingly coming from extending, restarting, and uprating the existing fleet, turning “end-of-life” plants into long-lived, high-output assets. In Illinois, the Clinton Power Station operated by Constellation had been effectively priced out by cheap shale gas, until Meta signed a 20-year agreement to fund a life extension in exchange for the carbon credits associated with nuclear generation.

Constellation CEO Joe Dominguez argues that the quickest “new” nuclear capacity will come from squeezing more out of the reactors already running, saying the existing U.S. fleet could add “7–10GW” through incremental upgrades and performance improvements. 

SMRs and “Next-Gen Nuclear"

The second reframe is on the technology side. Interest is surging in small modular reactors (SMRs) and advanced designs which proclaim fast, safe, and flexible deployment models. However, the outlook remains unclear for these firms. Yes, venture money is back, but the pathway to deployment is still fraught with risk.

Above: a rendering of an OKLO SMR (Small Modular Reactor) facility Source: OKLO

The WSJ reports that since 2021, venture capitalists have invested $2.5 billion in U.S. next-generation nuclear technology—after years when investment “hovered near zero.”A vivid example is Standard Nuclear, rebooted with $42 million led by Decisive Point, a rare move for investors who historically avoided heavily regulated, capital-intensive sectors. The WSJ frames the allure as coming from AI power demand and, to an extent, strategic competition with China.

The Economist puts hard numbers on the SMR funding wave: SMR startups alone raised another $2 billion since early 2024, including $460 million for Oklo and $650 million for TerraPower, and it notes Google’s deal to help Kairos Power develop an SMR fleet by 2035.

The technical and regulatory challenges remain: U.S. firms have not finished building an SMR. Meanwhile, critics argue that small reactors may not produce enough energy to justify their expense, and that startups underestimate the regulatory complexity involved to scale once they do reach maturity.

Policy Tailwinds—and a Regulatory Trade-Off

Finally, the political posture has turned sharply supportive. WSJ reports President Trump signing four executive orders aimed at accelerating nuclear deployment, and describes industry excitement that the orders could slash red tape—potentially allowing reactors on federal lands and “largely bypassing” the Nuclear Regulatory Commission.That enthusiasm is paired with an explicit warning: critics argue the push signals less NRC (Nuclear Regulatory Commission) independence and a more permissive environment for startups without a long safety record. The Economist adds a parallel message of caution: “excessive regulation is still a barrier,” but efforts to slash red tape may be too aggressive in an industry where the risk is as substantial as this one.

What’s Actually New Here

Taken together, these conditions describe an industry that is building a formula to tackle a massive technological, economic, and regulatory bear. Between reliable power demand from AI hyperscaling, long-term contracts, and political urgency, the nascent nuclear renaissance has some significant momentum. The near-term gains are likely to come from preserving and upgrading the existing fleet, but SMRs remain a high-upside bet that depends on converting first deployments into repeatable manufacturing and finance. And over all of it hangs the real kicker that past “renaissance” narratives have always downplayed: can you convert the dream you sold to investors into real power on the grid?

With the conceptual groundwork laid, the metric for success is now in poured concrete and delivered megawatt-hours, not term sheets and press releases.

Read More Tech Talks by Rob Cummins: The Wall-less Empire: The Illusion of Open AI Markets The SpaceX Diaspora: How Rocket Science Is Transforming Startup Culture