Sustainability & AI
AI energy consumption
The electricity used to train and run AI models, concentrated in data centres and growing with model scale.
Definition
The electricity consumed in developing and operating AI systems, concentrated in data centres and growing with model scale and usage.
Quick reference
At a glance
- Subject
- Sustainability & AI
- Editorial status
- Editorial draft
- Definition status
- Established
- Last updated
- 21 August 2026
References
This source provides part of the technical or institutional basis for the definition.
This source supports the explanation of how the term is applied, measured or governed in practice.
This source supports the wider sustainability significance and context described in the entry.
Overview
What it means
AI energy consumption includes both energy-intensive training runs and the continuing electricity used for inference. The IEA's estimate of about 415 TWh in 2024 covers all data centres—not AI alone. AI is a major growth driver in the IEA's central projection of roughly 945 TWh of total data-centre electricity demand by 2030.
How it is used
Grid operators now plan around data-centre load growth; providers compete to site facilities near low-carbon power; and regulators — the EU first — are imposing energy reporting on data centres and large AI models.
Why it matters
Electricity is the binding constraint on AI's expansion and the main channel of its climate impact. Where that power comes from — and how efficiently it is used — determines whether AI growth accelerates or undermines decarbonisation.
Review
Help keep this definition useful and accurate.
Submitted reviews stay private until an editor decides whether to accept and attribute them.
Endorse this definition
Confirm what works
Endorse the current wording when it is accurate and useful. Suggested changes use the separate editorial form below.