AI boom expected to drive global electricity demand to double by 2030

Global electricity consumption by data centres is projected to surpass 945 terawatt-hours (TWh) by 2030, more than doubling from 415 TWh in 2024, according to the International Energy Agency (IEA). This surge is largely driven by the rapid expansion of artificial intelligence (AI), especially generative AI, which demands extensive computing power and data processing capabilities.

At current trends, data centres could account for nearly 3% of global electricity usage by 2030, a level comparable to Japan’s entire national consumption. According to an article by Energy News, the Turkish-born Executive Director of the IEA, Fatih Birol, underscored the importance of proactive policy measures to mitigate rebound effects and manage the growing energy footprint of the digital sector.

Regional disparities in data centre infrastructure are significant. While a 100-megawatt facility already consumes as much power as 100,000 homes, emerging projects are scaling up to 2,000 megawatts—equivalent to the consumption of 2 million households. These mega-facilities are typically located near major cities, increasing strain on local electricity grids.

The United States, China, and Europe together represent 85% of global data centre electricity usage. Among these, the US is experiencing especially rapid growth, with data centres expected to drive nearly half of the country’s additional electricity demand in the coming years.

To address rising energy needs, the US has formed the National Energy Dominance Council under US President Donald Trump’s direction in 2025, aiming to expand electricity production and ensure energy competitiveness with China. The IEA notes that while coal currently supplies 30% of data centres’ energy, a transition to natural gas and renewables is likely, driven by cost and accessibility factors.

The environmental implications are considerable. CO₂ emissions from data centres are expected to climb from 180 million tonnes in 2024 to 300 million tonnes by 2035. However, this would still account for less than 1.5% of total emissions from the global energy sector. The IEA adds that some of the increased emissions may be offset by efficiency gains enabled by AI technologies.

Despite AI’s potential to support energy optimisation, the IEA cautions against overestimating its role in addressing climate challenges. Increased adoption of AI could be accompanied by higher fossil fuel consumption, limiting its net benefit. The agency stresses that no single technology, including AI, can independently solve the complex issues of the global energy transition.

By Nazrin Sadigova