Skeptics doubt new reactors, but AI demand, defense projects, and federal financing could spark a real US nuclear energy revival.
In the sea of enthusiasm for the global expansion of commercial nuclear power, there are modern-day Nostradamuses confidently predicting failure.
“Most [small modular reactors] SMRs are on paper [… and] the tech bros don’t seem to be grounded in reality,” prophesized a Former Nuclear Regulatory Commission chairperson in a recent BBC interview.
On social media, a noted nuclear power critic predicted, “you must know that none of this is actually going to happen” when referring to new nuclear microreactors “that are doomed to fail.”
It is very possible that these criticisms will be proven correct, and small reactors under 300 megawatts (MW) may lose steam as they slog through the thicket of design, licensing, and demonstration. Certainly, a new report on OKLO raises numerous concerns about technological readiness.
Also, as the critics note, there are no new commercial nuclear plants in the United States under construction or even under contract by energy utilities at this time. There are lots of memoranda of understanding (MOU) and paper reactors, but not much groundbreaking.
But the Trump administration has promulgated reams of requirements for advancing nuclear power, including adding 300 gigawatts (GW) of new nuclear energy capacity by 2050, and now has a plan to add financial fuel to this process.
The question is whether the critics are right and there are too many holes for projects to fall through, or whether the Trump mandates will shock the slow-to-change nuclear energy ecosystem into supercharging deployment.
AI Energy Expansion
The Trump administration’s nuclear energy objectives are fueled by several imperatives, but a core one is to win the artificial intelligence (AI) race with China.
That means it needs to concentrate capital and energy resources on the data center buildout. Today, the United States has 522 data centers, about half of the world’s total. Another 280 are expected to come online by 2028. This would jump energy demand in this sector from two percent of total US electricity to 12 percent.
The United States is currently adding only about 65 GW to the grid, but according to an ICF International analysis, it “should be adding about 80 gigawatts of new power generation capacity a year to keep pace with AI as well as cloud computing, crypto, industrial demand, and electrification trends.”
The last nuclear power capability the United States added was the completion of 2.2 GW at the Alvin W. Vogtle Electric Generating site in Georgia in 2024. By contrast, China has 33 reactors under construction and another 43 planned.
Since then, several shuttered nuclear plants have been slated for refurbishment in Pennsylvania, Michigan, and Iowa. They will contribute new energy in the next few years, along with power uprates for existing reactors that are expected to be approved by the Nuclear Regulatory Commission (NRC).
The partially completed VC Summer nuclear plant in South Carolina will be finished now that the Santee Cooper utility has chosen Brookfield, Westinghouse Electric Company’s majority owner, to complete the work on the AP-1000 reactors.
A number of tech companies, including Amazon, Google, and META, are investing in small reactors that will power the AI race, and Microsoft is supporting the restart of a shuttered reactor at Three Mile Island.
Reactor Role Reversal
The critics are betting that the challenges of achieving commercialization will sink the small reactor data center rally. Driving this view is a trifecta of concerns about emerging reactors’ cost-effectiveness, licensing ability, and delivery certainty. It may take a decade or more before these reactors are ready for deployment. By then, the AI boom may have slaked its energy thirst.
Evidence is already accumulating that proven large reactors in the one GW range, like Westinghouse’s AP-1000, are becoming more appealing for AI power after being cast aside in favor of the smaller newcomers.
Except for NuScale’s Voygr small modular reactor, no US-based SMRs have received NRC design approval, while the AP-1000 is licensed and has been built in the United States and abroad.
But AI is not the only application for small reactors. There are industrial applications for small reactors, including process heat, and small and disaggregated power grid uses that exist in developed and developing countries. The X-Energy deal with Dow to power the Seadrift plant is a good example of matching power plant output and consumption on one large industrial footprint.
Outside the United States, the Canadian Province of Ontario has approved the construction of four 300 MW General Electric Hitachi BWRX reactors at the Darlington nuclear plant. Ontario and the federal government committed $3 billion Canadian dollars for the project. If licensed, these would become the first commercial SMRs operating in North America. That will provide important concrete cost and construction information that may be applicable to other small reactor deployments.
Financing
Spinning up the American nuclear buildout will require billions in financing, and the history of significant cost overruns on previous plant construction casts a shadow on private sector financing enthusiasm.
The Trump administration, however, does not seem to be deterred by this history. It has embraced and revamped the Department of Energy’s Loan Program Office into the Energy Dominance Financing Program (EDF), which has an objective of “reinvigorating the nuclear industry.” It demands that projects move forward expeditiously.
EDF seems prepared to provide tens of billions of dollars in nuclear project loans beyond the commitments it has made to date. And, while the administration is willing to work with utility companies on this nuclear build-out, it is not waiting on them.
As part of its shockwave shake-up of the US nuclear industry, the Commerce Department has entered into a binding term sheet to facilitate financing of $80 billion for the domestic and overseas deployment of the AP-1000. The US government will be granted a participation interest of 20 percent of any cash distribution over $17.5 billion earned by Westinghouse if it secures the funding and assures the reactors will be completed.
Large reactors may be the next residents on federal land if Fermi America is any indication. That project aims to become the country’s largest nuclear complex, with multiple AP-1000s at the site.
But several factors need to fall together for the free flow of money to matter, including creating an assured fuel supply, developing significant efficiencies in the reactor Engineering, Procurement, and Construction (EPC) process to reduce overall cost and construction time, and rebuilding or importing a skilled labor cadre.
This is where US allies like South Korea could play an important role, particularly on AP-1000 construction. South Korea has experience with AP-1000s at home, has a world-class nuclear construction capability, produces components America can’t, and possesses a substantial and well-trained workforce. Korean companies are already partners in some US SMR projects.
Fuel Supply
Access to high-assay low-enriched uranium (HALEU) nuclear fuel is a major bottleneck for small reactors, and fuel supply could become a challenge for the US nuclear fleet if domestic uranium enrichment can’t quickly scale up.
Both the Biden and Trump administrations made revitalizing US nuclear fuel production a priority. The reasons were to reduce dependence on Russian fuel imports, increase domestic uranium enrichment, and ensure that a steady supply of fuel is available. Collectively, the two administrations have devoted billions to uranium enrichment.
The Trump administration is also enamored with plutonium as a fuel, and recently offered to provide nuclear companies access to almost 20 tons of weapon-grade plutonium for reactor fuel. Tens of millions of dollars in research on plutonium reprocessing technologies have been provided by both administrations.
Nuclear critics have cited nuclear proliferation problems associated with the use of HALEU and weapon-grade plutonium in the civil nuclear fuel cycle. For now, the administration is pushing forward on multiple nuclear fuel fronts, but it is not certain that its efforts will succeed in substantially breaking out of the reactor fuel bottleneck.
Military Reactors
While SMR skeptics are relying on the commercial requirements to smother the viability of next-gen reactors, they have not adequately accounted for the commitment of the US Department of Defense (DOD) to the development of small reactors. The key projects are Project Pele and the Janus Program. And DOD has vast experience turning nascent technology into commercial products.
The PELE reactor is a 1.5 MW TRISO-fueled, gas-cooled demonstration micro reactor that is manufactured by BWXT Advanced Technologies. The project broke ground at Idaho National Laboratory (INL) in 2024, and the reactor is expected to begin operation in 2028. The goal is to provide the US military with “a mobile, reliable, sustainable, and resilient power source which does not require a long logistics tail.”
The Janus project was launched in October 2025, and it is designed to be a micro reactor of one to 20 MWs that provides power for national defense facilities and critical missions. The program is a response to Trump’s executive order, which called for the operation of a US Army-regulated nuclear reactor at a domestic nuclear installation by September 2028.
The Janus reactors will be commercially owned and operated and could provide power to the local grid. The Army wants pairs of reactors demonstrated to underscore that the effort is not just research and development but aimed at wider-scale deployment.
There are questions about the ability of a reactor controlled by the defense department to provide commercial power, but the Atomic Energy Act does provide some authority.
Export Markets
The export of American nuclear reactors has slowed dramatically over the past 40 years but is now showing some signs of life. The AP-1000 is slated to be built in Poland, with the reactors possibly being deployed in Finland, Bulgaria, Slovenia, and Ukraine. Much of this interest is driven by the need to decouple from Russian energy.
US small reactors, however, are not faring as well in the export market in part because of the focus on the need to demonstrate them in the United States first.
However, these SMRs have applicability as an export technology, particularly to developing economy nations with growing populations and energy demand. But there is the need to first develop the nuclear power capacity and infrastcuture in many of these nations.
The United States has a slow, stepwise process for developing nuclear power capacity in countries that need assistance. It emphasizes training, technical assistance, and knowledge sharing. It has not resulted in the development of many markets for emerging US reactors, although it has provided SMR simulators to Romania and Ghana.
That process needs to evolve into one that identifies target countries and then applies the necessary technical and financial resources to develop them for the import of US small reactors. The think tank, Clearpath, has developed the concept of Energy Security Compacts (ESC) that would align US nuclear and financial tools to achieve this goal. It proposes 5 to 10-year agreements that align the various bureaucratic elements of the US government to work toward “clear, measurable outcomes” that support US foreign policy and energy security objectives.
This concept would strengthen the ability of the United States to win nuclear markets against the state-financed companies of Russia and China, which are already operating their own SMRs. And seizing markets outside of North America for US small reactors would push forward their commercialization in America, thereby responding to the concerns of the critics.
About the Author: Ken Luongo
Kenneth Luongo is the president and founder of the Partnership for Global Security (PGS) and the Center for a Secure Nuclear Future.
Image: Sergey Nivens/shutterstock
