IEA: Fears That AI Will Accelerate Climate Change “Appear Overstated”

A comprehensive new report from the International Energy Agency finds that concerns over artificial intelligence accelerating climate change are likely overstated. While data centers that power AI require increasing amounts of electricity — often from fossil fuels — the report highlights potential climate benefits from AI’s broader applications.

AI’s Growing Energy Demand

According to the IEA, data centers consumed around 1.5% of global electricity in 2023. That figure is projected to more than double by 2030, exceeding the current electricity use of Japan. A typical AI-dedicated data center already uses as much power as 100,000 households, and new mega-centers under construction could use 20 times more.

Despite this energy surge, the IEA’s base-case scenario shows renewables playing the largest role in meeting data center demand through 2035, followed by natural gas. However, grid strain is already an issue in some regions. The report warns that around 20% of planned data center projects may face delays unless infrastructure risks are resolved.

Emissions: Threat or Opportunity?

The IEA doesn’t frame AI as either a major climate threat or a miracle solution. Instead, the report presents a nuanced view, noting that AI has the potential to reduce emissions through various applications.

For example, AI can help improve grid efficiency, enhance renewable energy integration, and detect methane leaks in oil and gas operations. The report estimates that if existing AI solutions were widely deployed, they could cut emissions by the equivalent of 5% of energy-related emissions by 2035.

“The widespread adoption of existing AI applications could lead to emissions reductions that are far larger than emissions from data centres — but also far smaller than what is needed to address climate change,” the IEA concluded.

A Clouded Future

The IEA acknowledges uncertainty in its projections, offering three alternative models in addition to its base case. Variables include global regulations, technological innovation, and industry growth trends. The agency also notes that the word “bubble” is beginning to circulate when discussing AI’s trajectory, underscoring the unpredictability of both its benefits and its risks.

Bottom Line

While AI-powered data centers will contribute to rising electricity demand, the IEA suggests that their climate impact may be mitigated — or even outweighed — by emissions-cutting technologies that AI enables. Still, the long-term outcome depends on policy choices, energy mix evolution, and how quickly climate-positive AI applications are adopted.