THE DAY, IN CONTEXTPolaris InsightsAn AI-powered comprehensive news portal

POLARIS INSIGHTS

How Trump and AI Data Centers Are Boosting Nuclear Power — EnergyNow.com

Trump administration’s nuclear push, backed by AI data‑center demand, includes funding, regulatory changes and reactor restarts, while SMR rollout remains years away.

Nuclear power is seeing renewed interest in the United States as data‑center growth for artificial‑intelligence models drives demand for reliable, carbon‑free electricity. The Electric Power Research Institute projects data‑center electricity use could rise to 17% of total U.S. consumption by 2030, prompting tech firms to view nuclear as a 24/7 clean‑energy source. Public support for more nuclear has also risen, with 59% of adults favoring it in 2025, according to Pew Research Center.

The Trump administration is pursuing a “nuclear renaissance” by setting a goal to quadruple nuclear capacity to 400 GW by 2050 and to have at least ten large reactors under construction by decade’s end. It has pledged more than $80 billion for Westinghouse‑designed reactors, signed an executive order to halve Nuclear Regulatory Commission licensing timelines to 18 months, and directed the NRC to revisit radiation‑limit guidance. The Department of Energy’s reactor pilot program selected 11 projects, including two from Oklo, and aims for three advanced designs to reach criticality by July 4 2026; three startups met that milestone.

Several existing plants are slated for restart—Holtec plans to reopen Michigan’s Palisades plant, Microsoft has contracted for power from Three Mile Island, and NextEra seeks to revive Iowa’s Duane Arnold plant for Google. Small modular reactors (SMRs) are being developed, with NuScale receiving NRC design approval and TerraPower obtaining a construction permit, but no SMR has yet earned an operating license. Tech companies such as Meta have signed power‑purchase agreements with SMR developers, yet industry consensus holds that widespread SMR deployment is unlikely before the early 2030s, and cost and licensing uncertainties remain.

More from the AI desk

More from AI · Back to Polaris Insights