In April 2025 the International Energy Agency published “Energy and AI,” its first major report dedicated to the two-way relationship between artificial intelligence and the power system. The report opens from a blunt premise: there is no AI without energy, specifically electricity for data centres, and it sets out to quantify how large that demand is becoming.
The headline projection is that global data centre electricity consumption is set to more than double, from around 415 terawatt-hours in 2024, about 1.5 percent of the world’s electricity, to roughly 945 terawatt-hours by 2030. For scale, the IEA notes that the 2030 figure is slightly more than Japan’s total electricity consumption today. AI is identified as the single most important driver of this growth, with the United States accounting for the largest share of the projected increase, followed by China. The report also explores the other side of the ledger, arguing that AI can help the energy sector itself by improving grid management, forecasting, and efficiency.
The report turned the energy footprint of AI from an environmental side note into a central economic and strategic question. Whether enough firm, clean power can be built fast enough to feed AI data centres now shapes capital plans for hyperscalers, utility load forecasts, and national industrial policy, making electricity supply one of the genuine constraints on how quickly AI can scale.