The Trump administration plans to supercharge the deployment of nuclear power, according to the Wall Street Journal, as baseload power demand surges from AI data centers and broader grid electrification. Hyperscalers are set to spend roughly $800 billion on data-center capex this year alone, intensifying strain on an already stretched U.S. electric grid.
Intermittent solar and wind are increasingly viewed as insufficient to meet the scale and reliability requirements of the modern economy. Nuclear power is emerging as the clean, always-on alternative needed to power the AI era, with proponents arguing it can provide the steady baseload that variable renewables cannot guarantee.
Utilities across the country are scrambling to keep up as power-hungry AI centers expand at a staggering pace, often outpacing grid capacity. The reported federal push aims to accelerate permitting and deployment of new reactors, though specifics on timeline or funding remain unclear.
America's electric grid is entering a period of unprecedented strain, with data-center growth coinciding with reshoring and broader electrification efforts. The grid's vulnerability to space weather events adds another layer of risk, as geomagnetic storms can disrupt transformers and cause blackouts, further threatening AI infrastructure.
The nuclear gambit faces a significant counter-argument: critics question whether regulators and the industry can scale new reactor designs quickly enough to meet near-term demand, given decades-long construction timelines and unresolved waste storage issues.