Datacentres to create 70% more carbon emissions by 2030, IEA report warns

The Guardian
Read full story· 4 min
The International Energy Agency warns datacentres will emit 70% more carbon by 2030, driven by AI growth and reliance on fossil fuels in the US and China. AI datacentres will double electricity use, with China relying heavily on coal and the US on gas.

What happened

The International Energy Agency (IEA) released a report stating that global datacentres are projected to produce 70% more carbon emissions by 2030. This increase is attributed to the rapid expansion of artificial intelligence and continued dependence on fossil fuels, particularly in the United States and China. The report highlights that AI datacentres will use twice as much electricity as they do today, with China relying mainly on coal and the US on gas to meet this demand. The findings are based on recent data about AI server shipments and the construction of large-scale AI datacentres worldwide.

Key details

  • AI datacentres are expected to consume an extra 350 terawatt-hours of electricity annually by 2030.
  • About half of China's AI datacentre electricity will come from coal, while 60% of US datacentre electricity will come from gas.
  • The US has plans for 180GW of new gas-fired power plants linked to datacentres, nearly equal to the current gas capacity of the EU.
  • Datacentre capacity has increased tenfold in the last two years, reaching 17GW, with 80% of current capacity in China, the US, and the EU.
  • The EU is expected to supply about 70% of its datacentre electricity from renewables by 2030.

Why it matters

Companies operating datacentres, especially in the US and China, will face higher energy costs and increased scrutiny over carbon emissions as they expand AI infrastructure. Policymakers and energy providers in these regions must address the growing demand for electricity and the environmental impact of increased fossil fuel use. The shift in energy sourcing for datacentres will also affect energy markets and the pace of renewable energy adoption in different regions.


More in Energy & Climate