AI Infrastructure

AI Data Centers and the U.S. Power Grid’s “Watershed Moment”: Electricity Supply Is Becoming the Next Constraint for the AI Industry

The Belfer Center for Science and International Affairs at Harvard Kennedy School and the Electricity and AI Initiative at the Harvard School of Engineering and Applied Sciences jointly released a policy brief, "AI, Data Centers, and the U.S. Electric Grid: A Watershed Moment." The brief notes that, driven by hyperscale data center development, the growth rate of U.S. electricity consumption is rising again, and Lawrence Berkeley National Laboratory predicts that data center demand will increase from 176 terawatt-hours in 2023 (4.4% of national electricity use) to 325–580 terawatt-hours by 2028 (6.7%–12.0%). In some regions, AI-driven electricity demand has already exceeded available capacity, forcing companies to delay projects, buy power directly from private generators, or install multiple less-efficient natural gas reciprocating generator sets. The brief also flags the risk of stranded assets, disputes over cost allocation, and the possibility of regulatory intervention, and uses Virginia and Texas, the two largest data center markets, as examples to show differing regulatory tempos. Its core conclusion is that excessive regulation could slow AI development, while insufficient regulation could lead to grid instability, higher residential electricity prices, and setbacks to climate goals.