The U.S. government commits $17.5 billion in funding to accelerate nuclear energy development and SMR deployment.

The Department of Energy has unveiled up to $17.5 billion in conditional loan commitments to help utilities purchase long-lead components for new nuclear reactors using Westinghouse's AP1000 technology. The program is explicitly tied to data-center electricity demand, marking perhaps the clearest federal signal yet that advanced nuclear is viewed as a critical solution to powering the AI infrastructure build-out.

While the specific utilities participating in the program have not been named, the structure of the loans—targeting components for Westinghouse AP1000 reactors—points to a narrow set of likely beneficiaries.

**Cameco** is probably the most direct play on the Westinghouse reactor fleet. The Canadian uranium producer co-owns Westinghouse with Brookfield Renewable Partners, giving it exposure to both fuel supply and the reactor technology the loan program is designed to support. Each new AP1000 built through these loans could drive long-term demand for Cameco's uranium business and its Westinghouse stake. Cameco's position is unique among likely beneficiaries because it operates on both sides of the transaction—supplying uranium and fuel services while owning a stake in the company designing and procuring equipment for new reactors. The main risk is that the loan program is conditional and nuclear build timelines are measured in years, not quarters. Near-term earnings are unlikely to be affected, though the long-term demand picture could improve materially if a significant portion of the $17.5 billion is deployed.

**Constellation Energy**, the largest nuclear fleet operator in the United States, has been the most visible utility in the data-center power narrative. Its existing nuclear plants are already being monetized through power purchase agreements with hyperscalers, including a high-profile deal to supply Microsoft's data-center operations. The DOE loan program could eventually extend this thesis to new builds, potentially locking in years of contracted revenue at premium prices. However, execution remains a major risk. New nuclear builds in the United States have a poor track record; the only AP1000s completed in the past decade, at Vogtle in Georgia, ran years late and billions over budget. Constellation's willingness to accept that risk will depend on the structure of federal support and customers' willingness to sign long-term contracts at prices that justify the build.

**Vistra**, the Texas-based power producer, rounds out the list. The company has built one of the largest clean-generation fleets in the country, including a meaningful nuclear footprint acquired through its merger with Energy Harbor. Vistra's stock has been among the best performers in the power sector over the past two years, driven by the same data-center thesis underlying the DOE program. The company offers a combination of contracted nuclear capacity, a fast-growing retail business, and a balance sheet capable of taking on new power generation if the economics work. Though Vistra has been more cautious than Constellation about committing to new nuclear builds, the DOE loan program could shift that calculus by reducing upfront capital risk.

It is important to note that the $17.5 billion is a conditional loan commitment, not a disbursement. Utilities must apply, qualify, and commit to building before any funds are deployed. The structure is designed to derisk long-lead component purchases, which have historically been a bottleneck for new nuclear development in the United States. For the three companies above, the program represents a potential long-term tailwind rather than a near-term catalyst. More direct catalysts—signed offtake agreements with data-center customers, construction start dates, and final loan closings—remain ahead.