The rapid expansion of artificial intelligence is creating new pressure on electricity systems, prompting warnings from U.S. energy officials about potential power shortages. Speaking at an industry discussion, U.S. Energy Secretary Chris Wright said the explosive growth of AI data centers could strain power infrastructure if energy supply does not expand quickly enough. AI systems require massive computing power, and the data centers that run them consume enormous amounts of electricity.
According to experts, modern AI data centers operate at extremely high power levels. Some of the largest facilities require more than one gigawatt of continuous electricity, which is roughly enough to supply hundreds of thousands of homes. As tech companies race to build new data centers to train and run advanced AI models, electricity demand is rising rapidly, placing pressure on power grids and energy infrastructure.
Energy analysts say this surge in demand could push overall electricity consumption to record levels. Forecasts from the U.S. Energy Information Administration suggest that national power use could reach 4,260 billion kilowatt-hours in 2026 and 4,388 billion kWh in 2027, driven largely by AI data centers, cryptocurrency mining, and broader electrification of homes and businesses.
Despite the concerns, officials also believe that increased energy demand from AI could encourage greater investment in power generation and infrastructure. Wright argued that states attracting energy investments for data centers may benefit economically, but warned that building the necessary facilities will require skilled labor and large-scale infrastructure projects. Without sufficient expansion of the energy grid, the AI boom could face serious limitations due to electricity shortages.
