US moves to expand nuclear power as AI data-center demand surges; NRC streamlines licensing
The US Nuclear Regulatory Commission is streamlining licensing pathways for advanced reactors to meet surging AI-driven electricity demand. NRC Chair Ho K. Nieh confirmed the agency has reformed its regulatory framework, issuing a new technology-neutral licensing framework in March designed to accelerate commercial deployment of next-generation reactor designs. The push reflects White House backing for nuclear expansion as data centers—which consume electricity equivalent to 100,000 households each—strain the grid.
Tech giants including Microsoft, Google, Amazon, and Meta are signing nuclear power agreements directly; Goldman Sachs estimates data-center electricity demand could rise 160% by 2030. The International Energy Agency projects data-center consumption will grow from just over 3% of global electricity in 2023 to over 9% by 2030. Bloomberg NEF and the Department of Energy both project US data-center demand reaching 106 gigawatts by 2035, requiring both existing reactor restarts (like Three Mile Island) and new construction.
For AI operators and energy infrastructure investors, this signals a policy tailwind: nuclear is becoming the marginal baseload provider for hyperscale compute. Utilities and SMR developers face a 2031–2035 window for deployment. The structural tailwind is structural, not cyclical—utility capex and nuclear supply chains will reorient.
Sources
- Primary source
- StartupHub.ai
“Ho K. Nieh, Chairman of the U.S. Nuclear Regulatory Commission (NRC), recently discussed the progress and the agency's role in facilitating this expansion”
- Carbon Credits
“Goldman Sachs estimates that data center electricity demand could rise by 160% by 2030”