AI Data Center Growth Strains Global Energy Grids

The rapid adoption of artificial intelligence is triggering a global surge in electricity demand that threatens to outpace existing grid capacities. In the United States, the U.S. Department of Energy projects that data center energy use could double or triple by 2028, potentially consuming 12 percent of the total national electricity supply.

Regional pressure is particularly acute in Pennsylvania, where the Public Utility Commission predicts industrial power demand will grow by an average of 18.56 percent annually through 2030. PPL Corporation reports a data center pipeline of 31.8 GW in advanced planning, with growth in the PPL Electric Utilities service territory expected to exceed 51 percent per year. While some officials warn that these loads fundamentally change the energy landscape, Joe Helfrich of NosTerra Venture Capital argues the pressure creates a unique opportunity to accelerate innovation in clean energy and storage.

Similar trends are appearing in China, where Wood Mackenzie projects data center electricity consumption will nearly quadruple by 2030 to reach 774 terawatt-hours. This growth is driven by intensive AI training and inference applications. Data from the National Energy Administration shows that electricity used by internet data services already surged 44 percent year-on-year in the first half of 2026. To manage this, the Chinese government is using its 15th Five-Year Plan to shift computing capacity toward renewable-rich western regions, with industry carbon emissions expected to peak by 2035.