AI Could Require Power Equivalent to 50 Large Nuclear Plants by 2030

Artificial intelligence is poised to significantly increase U.S. peak electricity demand, potentially reaching 50 gigawatts by 2030, according to a new analysis by the Electric Power Research Institute (EPRI). This figure exceeds half of the energy capacity of all U.S. nuclear plants, underscoring the growing strain AI could place on the nation’s power grid.
The report highlights the energy-intensive nature of AI training, where computer models process vast amounts of data to identify patterns and relationships. This process already accounts for significant electricity consumption by major tech companies like OpenAI, Meta, Amazon, Google, and Microsoft. As AI adoption accelerates, these demands are expected to rise sharply, creating challenges for grid operators.
Additionally, President Donald Trump’s AI strategy, aimed at securing U.S. leadership in the technology, could further complicate energy demands. The strategy encourages the development of large-scale data centers, potentially on federal land, which could strain regional power supplies and require new infrastructure investments.
As AI continues to expand, addressing its energy demands will be critical to ensuring a sustainable and reliable power grid for the future.

Published: 8/11/2025