Can Small Modular Reactors Deliver Power for AI Data Centers? July 26, 2026 Small modular reactors (SMRs) are frequently discussed as a potential long-term solution for powering AI data centers, but significant regulatory and deployment hurdles remain before they can play a meaningful role. While interest from hyperscalers is high, the first commercial SMRs in the United States are still years away from operation, with most realistic timelines pointing to the early 2030s rather than the late 2020s. The Regulatory Framework The U.S. Nuclear Regulatory Commission (NRC) oversees the licensing of all commercial nuclear reactors, including SMRs. Historically, the NRC’s licensing process was designed for large, traditional light-water reactors. This created challenges for smaller, advanced designs like SMRs, which often use different technologies and safety approaches. In response to growing electricity demand from data centers and AI, the federal government has pushed for faster regulatory pathways: A May 2025 executive order directed the NRC to complete reviews of new reactor designs within a maximum of 18 months . The NRC is developing Part 53 , a new technology-inclusive, risk-informed regulatory framework better suited for advanced reactors. It is expected to be finalized by late 2027 or earlier. In 2026, the NRC proposed Part 57 , a new framework specifically for microreactors, aiming to reduce licensing timelines to 6–12 months for qualifying designs. These reforms represent the most significant effort in decades to modernize nuclear regulation in the United States. Current Status of SMR Deployment Despite regulatory progress, actual deployment remains limited: NuScale Power is currently the only company with full NRC design certification for an SMR (its 77 MWe design was approved in May 2025). As of mid-2026, no SMRs are under construction in the United States. Many projects originally targeted for the late 2020s have seen their timelines slip into the early 2030s. Major tech companies have shown strong interest in nuclear power. Microsoft has a deal to restart a unit at the retired Three Mile Island plant, while Amazon and Google are exploring small modular reactors and nuclear restarts. However, these projects are still in early stages. Why Timelines Matter for AI AI data center power demand is growing rapidly and is expected to increase significantly over the next decade. Many companies are seeking reliable, carbon-free power sources that can be deployed relatively quickly. SMRs offer theoretical advantages for data centers, including smaller size, factory fabrication, and the potential for co-location. However, the long lead times for regulatory approval, construction, and fuel supply mean SMRs are unlikely to provide meaningful capacity in the near term (2026–2029). In the short term, most behind-the-meter power solutions for data centers continue to rely on natural gas turbines and fuel cells, which can be deployed much faster than nuclear. The Path Forward Regulatory reforms are improving the outlook for SMRs, but several challenges remain: Finalizing and implementing new rules like Part 53 and Part 57. Completing first-of-a-kind projects to demonstrate real-world performance and costs. Navigating state-level regulations and public acceptance. If these reforms are successfully implemented and early projects succeed, SMRs could become an important part of the long-term power mix for AI infrastructure. For the next several years, however, their contribution is expected to remain limited. Sources U.S. Nuclear Regulatory Commission announcements on Part 53 and Part 57 (2026) Reporting from Reuters, The Bulletin of the Atomic Scientists, and industry sources on SMR timelines and deployment status Company announcements regarding nuclear deals by Microsoft, Amazon, and Google This article examines the current regulatory landscape and deployment outlook for small modular reactors in the context of AI data center power demand.