Worried about the carbon footprint of AI? The IEA’s April 2025 report sees data center electricity use more than doubling to around 945 terawatt-hours by 2030, yet the added emissions look small in the wider energy picture.
This guide sits within our wider look at the environmental impact of technology, and it focuses on the electricity side of AI.
| Item | Value |
|---|---|
| Global data center electricity demand in 2030 (IEA projection) | Around 945 terawatt-hours |
| Change in data center electricity demand by 2030 | More than double |
| Electricity demand of AI-optimized data centers by 2030 | More than quadruple |
| US data centers’ share of US electricity demand growth to 2030 | Almost half |
| Data centers’ share of demand growth in advanced economies to 2030 | More than 20% |
These figures come from the American Public Power Association’s summary of the IEA’s Energy and AI report, published in April 2025. Projections change as new data arrives, so check the source before you rely on them.
Why electricity is at the heart of AI’s footprint
AI runs in data centers, and data centers run on electricity. So the carbon footprint of AI depends on how much power those buildings draw and on how that power is generated.
The IEA’s special report Energy and AI, released on April 10, 2025, is a useful starting point. It draws on new datasets and extensive consultation with policy makers, the tech sector, the energy industry and international experts, as reported by the American Public Power Association.
It helps to keep apart the questions that often get mixed together:
- How much electricity AI and data centers will use.
- Which energy sources will supply that electricity.
- What AI might save elsewhere in the energy system.
How much electricity will data centers use by 2030?
The IEA projects that electricity demand from data centers worldwide will more than double by 2030, to around 945 terawatt-hours. The figure comes from the APPA summary of the report.
A terawatt-hour is a unit of energy used for national and global totals, not for single devices. That is why it appears in forecasts like this one.
AI is described as the most significant driver of the increase. Electricity demand from AI-optimized data centers is projected to more than quadruple by 2030, so the growth is not spread evenly across all computing.
Where will the extra demand land?
In the United States, power consumption by data centers is on course to account for almost half of the growth in electricity demand between now and 2030, as of the IEA’s April 2025 report. The APPA article adds that, driven by AI use, the US economy is set to consume more electricity in 2030 for processing data than for manufacturing all energy-intensive goods combined, including aluminum, steel, cement and chemicals.
In advanced economies more broadly, data centers are projected to drive more than 20% of the growth in electricity demand over the same period. According to the same source, that puts the power sector in those economies back on a growth footing after years of stagnating or declining demand in many of them.

Power sources and the emissions picture
A diverse range of energy sources is expected to meet data centers’ rising needs. As of the IEA’s April 2025 report, renewables and natural gas are set to take the lead because of their cost-competitiveness and availability in key markets, per the APPA summary.
The report says higher electricity demand for data centers is set to drive up emissions. It also says this increase will be small in the context of the overall energy sector. It could potentially be offset by emissions reductions enabled by AI if adoption of the technology is widespread.
That last point is a conditional, not a promise. The same report finds AI could accelerate innovation in energy technologies such as batteries and solar PV, but the savings depend on how widely and how well AI is used.
Beyond electricity: minerals, equipment and security
Power use is not the only pressure. The IEA report provides first-of-its-kind estimates of data center demand for critical minerals, whose global supply is today highly concentrated, according to the APPA article.
Hardware that reaches the end of its life is a separate issue, and we cover it in our explainer on what is e-waste.
Security also enters the picture. The IEA said cyberattacks on energy utilities have tripled in the past four years and become more sophisticated because of AI. At the same time, AI is becoming a critical tool for energy companies to defend against such attacks, as reported here.
If you run a website and want to look at the hosting side of your own footprint, our guide on what is green web hosting explains one option to consider.
What is still uncertain, and what to watch
The report stresses that significant uncertainties remain. They range from the macroeconomic outlook to how quickly AI will be adopted.
It also raises open questions, which are worth treating as a checklist when you read any forecast:
- How capable and productive AI will become.
- How fast efficiency improvements will arrive.
- Whether bottlenecks in the energy sector can be resolved.
For countries that want to benefit from AI, the report says they need to quickly accelerate new investment in electricity generation and grids. They also need to improve the efficiency and flexibility of data centers and strengthen the dialogue between policy makers, the tech sector and the energy industry, per the APPA summary.
As of April 2025, the IEA planned to launch an Observatory on Energy, AI and Data Centers to gather recent worldwide data on AI’s electricity needs. That report built on the Global Conference on Energy and AI the agency hosted in December 2024, so newer figures may have appeared since.
Frequently asked questions
How much electricity will data centers use by 2030?
Show answer
As of the IEA’s April 2025 report, global data center demand is set to more than double to around 945 terawatt-hours by 2030. Check the IEA for newer estimates before you quote it.
Is AI the main reason data center demand is rising?
Show answer
The report calls AI the most significant driver of the increase. Electricity demand from AI-optimized data centers is projected to more than quadruple by 2030.
Will AI’s emissions be large compared with the whole energy sector?
Show answer
According to the IEA report, the emissions increase from data centers will be small in the context of the overall energy sector. It could potentially be offset by AI-enabled reductions if adoption is widespread.
Which energy sources are expected to power data centers?
Show answer
A diverse range of sources will be used, but as of the IEA’s April 2025 report, renewables and natural gas are set to take the lead because of cost-competitiveness and availability in key markets.



