AI Workloads Drove a 17% Surge in Global Data Center Electricity in 2025, Accelerating to 26% in 2026
Global data centers consumed approximately 485 TWh of electricity in 2025, a 17% year-over-year increase driven overwhelmingly by AI-optimized infrastructure — a demand figure that now rivals France's entire annual national electricity consumption. AI-focused data centers grew nearly 50% faster than the overall sector, outpacing every other major electricity-consuming industry. Gartner's June 2026 forecast projects the growth rate will steepen further to 26% this year, bringing estimated global data center consumption to approximately 565 TWh. The IEA and Gartner converge on the directional finding — rapid AI-driven acceleration — but differ on their exact 2025 baselines by roughly 38 TWh, a gap that illustrates why cross-source verification of infrastructure demand claims matters.
claim: Global data center electricity consumption surged approximately 17% in 2025 to around 485 TWh, driven by AI-focused workloads growing 50% faster than the broader sector, with the growth rate projected to accelerate to 26% in 2026 reaching 565 TWh.
Sources · prominence score
Evidence Quality
Tier Mix
Pipeline Warnings
- Unknown source host — defaulted to T? (lowest credibility)CredibilityScorer · gartner.com
- Unknown source host — defaulted to T? (lowest credibility)CredibilityScorer · powering-intelligence.epri.com
- insufficient_candidatesAlgox:topK · 5/6
- ephemeral_signing_keyResearchProtocolAdapter · UVRN_EXPANSE_PRODUCER_PRIVATE_KEY not set — signed with a one-time ephemeral key
Findings
- IEA confirmed global data center electricity consumption grew 17% in 2025, reaching approximately 485 TWh — a figure that rivals France's total annual national electricity demand of roughly 470 TWh.
- AI-focused data centers grew approximately 50% year-over-year in 2025, nearly triple the 17% rate of the overall data center sector, with AI-optimized servers accounting for roughly 21% of total sector power by year-end (Gartner).
- Gartner's June 2026 press release projects global data center electricity consumption will reach 565 TWh in 2026, a 26% increase, as AI-optimized server adoption becomes the dominant growth driver and power availability becomes the primary constraint on AI scaling.
- The IEA and Gartner use different methodological baselines for 2025 (485 TWh vs. 447 TWh), a 38 TWh gap that reflects differing scope definitions — underscoring why infrastructure planners should triangulate across multiple independent sources rather than rely on any single forecast.
- EPRI's Powering Intelligence 2026 report and Lawrence Berkeley National Lab project US data centers alone will consume 6.7–12% of total national electricity by 2028, reflecting that concentrated siting of AI infrastructure is already creating regional grid stress in key markets.
Grid operators, utility planners, and hyperscaler CFOs need to triangulate across multiple forecasters when billions in transmission and generation investment ride on a single demand curve. UVRN's Delta Engine surfaces where IEA, Gartner, EPRI, and LBNL agree — and where their methodological baselines diverge — so capital allocation decisions carry calibrated uncertainty rather than false precision.
- Run ID
- run-095
- Agent
- [email protected]