AI Energy Consumption Index: Q2 2025 Edition

Global data centre electricity reached an estimated 485 TWh in 2025, with AI workloads driving 20–25% of that total and growing 50% year-on-year. This edition of the AI Energy Consumption Index sets out transparent, evidence-based estimates for the leading AI models and services, benchmarks UK demand against international peers, and projects the electricity trajectory to 2035 under three scenarios.

• Global data centre electricity consumption is on track to roughly double from 485 TWh in 2025 to 950 TWh in 2030; AI-focused data centres accounted for around a third of the 2025 total. IEA, Key Questions on Energy and AI (2026).
• Under the IEA Base Case, global data centre demand could reach 945–950 TWh by 2030 — roughly Japan's annual consumption. IEA, Energy and AI (April 2025).
• ChatGPT receives around 2.5 billion prompts per day from users worldwide, according to OpenAI. TechCrunch, 21 July 2025, citing OpenAI.
• Inference has overtaken training as the dominant driver of AI electricity use. Source needed
• UK data centres used around 5 TWh of electricity in 2023 (2% of total UK demand); Oxford Economics projects this reaching 26.2 TWh by 2030 (8.8% of total UK demand). Oxford Economics, The UK's data centre boom.
• Ofgem reports total contracted demand in the GB grid connection queue grew from 41 GW (November 2024) to 125 GW (June 2025) — a measure of contracted connection capacity, not electricity consumed, not exclusively data-centre demand, and not a guarantee every queued project connects; Ofgem itself has flagged concern that a significant share may not proceed. Ofgem proposed data centre connection reforms consultation Separately, NESO's connections reform restructured the broader UK-wide connections queue (covering both demand and generation projects): roughly 217 GW of projects lacking consents or land rights were removed, cutting that total from over 700 GW to around 238 GW under a new readiness-based system, with an estimated 13 GW of demand connecting before 2030 and a further 86 GW between 2030 and 2035. These are two distinct measures on different scopes and should not be read as the same figure before and after reform. NESO connections reform results

• Policymakers, regulators and DESNZ / Ofgem analysts
• Data centre operators, hyperscalers and grid planners
• AI companies, cloud providers and semiconductor investors
• Sustainability, ESG and net-zero leads at UK enterprises
• Journalists and researchers covering AI's energy footprint

1. 011. Executive summary and key findings — Q2 2025
2. 022. Methodology — how AI electricity estimates are calculated
3. 033. Executive dashboard — regional demand and service benchmarks
4. 044. State of AI — adoption timeline and market structure
5. 055. Electricity profiles — ChatGPT, Claude, Gemini, Copilot, Meta AI, Perplexity, Grok, DeepSeek
6. 066. Image generation — Midjourney and Stable Diffusion
7. 077. Model comparison — efficiency by class, open-source vs proprietary
8. 088. Data centres — AI server deployments, cooling and grid impacts
9. 099. Electricity trends 2020–2025 and quarterly peak demand
10. 1010. Water consumption — cooling systems and regional impacts
11. 1111. Carbon footprint — Scope 2, renewables and net-zero commitments
12. 1212. International comparison — UK, US, EU, China, Japan, Singapore, Middle East
13. 1313. Electricity demand forecast to 2035 — three scenarios
14. 1414. Industry analysis — semiconductors, nuclear, renewables and grid flexibility
15. 1515. Key risks and opportunities
16. 1616. Recommendations — government, energy companies, AI developers, cloud providers and investors
17. 1717. Glossary, references and detailed methodology appendices