AI Data Centers Consumed 415 TWh in 2024 — Fastest Electricity Demand Growth in Over a Decade
Global data centers consumed approximately 415 terawatt-hours of electricity in 2024, a figure that has nearly doubled since 2017 and now rivals the annual power consumption of Italy or Australia. AI workloads — model training, fine-tuning, and large-scale inference — are the primary driver of this acceleration, according to the International Energy Agency. Independent analyses from Lawrence Berkeley National Laboratory, Goldman Sachs Research, Wood Mackenzie, and Bloomberg all confirm the same directional picture: AI-fueled demand is growing faster than grid infrastructure can reliably accommodate.
claim: Global data centers consumed approximately 415 TWh of electricity in 2024 — nearly doubling since 2017 — with AI workloads driving the fastest demand acceleration in over a decade.
Sources · prominence score
Evidence Quality
Tier Mix
Pipeline Warnings
- Unknown source host — defaulted to T? (lowest credibility)CredibilityScorer · goldmansachs.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 reported global data centers consumed ~415 TWh in 2024, up from ~240 TWh in 2017 — roughly 73% growth in seven years, the steepest sustained rise since large-scale cloud computing emerged.
- Lawrence Berkeley National Lab found U.S. data centers alone consumed 176 TWh in 2023, with 2028 projections ranging 325–580 TWh depending on AI adoption pace — a potential tripling within five years.
- Goldman Sachs projects U.S. data center power demand reaching 41 GW by 2026, a 32% increase from 31 GW in 2025, with AI infrastructure buildout cited as the sole driver of incremental demand growth.
- Grid constraints are emerging as the binding bottleneck: Wood Mackenzie reported U.S. capacity additions slowed sharply in Q4 2025 due to interconnection backlogs and power scarcity, signaling that demand is outpacing the grid build-out.
Energy planners, hyperscale operators, and infrastructure investors use UVRN to verify whether AI power-demand projections from financial analysts align with government research and grid operators — critical for capital allocation, grid interconnection decisions, and site selection.
- Run ID
- run-118
- Agent
- [email protected]