AI Data Centers Push U.S. Electricity Demand Toward New Records in 2026 and 2027
The EIA projects U.S. electricity consumption will rise to 4,268 billion kWh in 2026 and 4,391 billion kWh in 2027 as AI and cryptocurrency data centers contribute to record power demand.
Why it matters: National electricity-demand forecasts quantify how AI infrastructure is reshaping the power market and turning grid capacity into a core technology constraint.
U.S. electricity consumption is heading toward consecutive record highs as artificial-intelligence data centers become an increasingly important source of power demand. The U.S. Energy Information Administration projects electricity use will rise from a record 4,195 billion kilowatt-hours in 2025 to approximately 4,268 billion kWh in 2026 and 4,391 billion kWh in 2027. AI and cryptocurrency data centers are among the major drivers, alongside broader electrification across the economy. The forecast reinforces one of the most important changes taking place in technology infrastructure: access to electricity is becoming nearly as strategically important as access to computing hardware. Large AI campuses can require hundreds of megawatts of power, forcing developers to compete for utility interconnections, generation capacity, transformers, substations and transmission infrastructure. EIA has estimated that data-center servers accounted for roughly 7% of U.S. commercial-sector electricity consumption in 2025. The latest forecast also shows how quickly infrastructure constraints can alter growth expectations, with EIA cutting its forecast for Texas electricity-demand growth in 2027 after a pause on new data-center development. The implication for the technology industry is increasingly clear: companies able to secure reliable power, grid access and physical infrastructure may gain an important advantage as AI computing continues scaling.