The AI boom is turning electricity into a strategic prize. In an IEA update published on April 16, 2026, the agency said global electricity demand from data centres jumped 17% in 2025, while overall electricity demand rose just 3%. AI-focused facilities expanded even faster. At the same time, the capital expenditure of five large technology companies exceeded $400 billion in 2025 and is projected to rise another 75% in 2026. The implication is striking: the race to dominate artificial intelligence is no longer only a contest for chips, talent, and capital, but also for access to reliable power. (iea.org)
Yet this is not a simple story of endless expansion. The IEA warns that AI is now colliding with hard physical limits: shortages of gas turbines, transformers, advanced chips, and grid-connection approvals are slowing projects. North America offers a vivid example. NERC’s 2025 Long-Term Reliability Assessment says new data centres tied to AI and the digital economy account for most of the projected increase in North American electricity demand over the next decade, with especially steep growth in Texas, PJM, and parts of the western grid. PJM’s own long-term forecast, posted on January 24, 2025, had to adjust multiple utility zones for data-centre growth and raised its 2030 summer peak forecast by 16,010 megawatts compared with the previous year’s report. In other words, the data-centre surge is no longer a niche planning issue; it is beginning to reshape grid governance itself. (iea.org)
This scramble for power also complicates decarbonisation. The IEA estimates that data centres used about 415 TWh in 2024 and could reach roughly 945 TWh by 2030, just under 3% of global electricity use. Renewables are still the fastest-growing source of supply for data centres, increasing at an average annual rate of 22% from 2024 to 2030 and meeting nearly half of demand growth. But that is only half the picture: natural gas and coal together are still expected to supply more than 40% of the additional electricity demand from data centres through 2030. The result is a genuine paradox. AI may accelerate efficiency elsewhere in the economy, but unless grids, transmission, and clean generation scale far faster, the age of intelligent machines could also prolong the life of carbon-intensive power. (iea.org)










