Energy and AI →
Examines electricity demand from AI and applications of AI in the energy sector. Global scenarios are not site forecasts.
RESEARCH & INTELLIGENCE
Evidence before
assumptions.
Selected official research and product announcements, reviewed on 12 November 2026. A source’s publication date is distinct from our review date.
415 TWhIEA estimate for global data centres in 2024.
945 TWhIEA Base Case projection for 2030.
Read the boundary
All data centres, not AI alone. A 2025 scenario, not a current measured total. Location, equipment efficiency and adoption can change outcomes.
Primary source reading room
Energy and AI: executive summary →
Estimates data centre electricity use at 415 TWh in 2024 and projects around 945 TWh in 2030 in its Base Case. Includes all data centres, not AI alone.
One Digital Grid Platform availability →
Describes an AI-enabled platform connecting utility planning, operations and asset management. Product statements are the vendor’s claims.
Gridscale X: system operations and agentic transmission planning →
Describes a shared grid model and AI-powered transmission planning. This is an announced vendor capability, not an independent performance assessment.
GridOS for Transmission and AI whitepapers →
Announces software for coordinated transmission operations and AI papers on grid planning and grid-edge autonomy.
Electrification X →
Describes an IoT suite for asset, load, microgrid and energy management across electrification infrastructure.
GridOS portfolio →
Describes grid orchestration software and applications. No vendor performance percentages are reproduced here.
Editorial method
Official publications establish what an organisation reported; they do not independently validate its performance claims. We distinguish estimates, projections, announced products and our own proposed applications. No market feed or automated refresh is connected.
We have intentionally omitted vendor return-on-investment statistics, live prices and unsupported market-size claims.
Download source ledger (JSON) ↓