[REVISION HISTORY]

AI infrastructure expansion: energy, grid, and resource use

Updated 49 times since CLSTR started tracking revisions of this situation.

What changed

2026-09-11 12:55 UTC → 2026-09-11 18:06 UTC · added ~~removed~~

AI infrastructure expansion: energy, ~~grid stability,~~ grid, and ~~new~~ resource use

AI-driven workloads continue to drive a rapid expansion of data-center capacity, fueling an “energy supercycle” characterized by sustained investment in power generation, transmission, and storage. Global electricity consumption by data centers is projected to nearly double from 485 TWh in 2025 to 950 TWh by ~~2030, representing~~ 2030. In the United States, electricity consumption by data centers was approximately ~~3% of global demand.~~ 4.4% in 2023 and is projected to triple by 2028. This surge could potentially drive wholesale electricity prices up by 50% by 2028. A critical mismatch persists between the speed of data center ~~construction—often operational within two to five years—and~~ construction and the decade-long timelines required for ~~upgrading high-voltage transmission lines and~~ grid ~~assets. The UN Economic Commission for Europe (UNECE)~~ upgrades. This has ~~warned that this lag poses risks~~ led to ~~grid stability, including potential voltage oscillations and unplanned outages. In response, regions such as~~ regional connection limits in Ireland and the ~~Netherlands have implemented connection limits~~ Netherlands, and ~~location constraints~~ growing public opposition in the U.S., where a Heatmap survey found 75% of Americans oppose new construction due to ~~prevent network overloading. To mitigate these pressures, companies are exploring diverse solutions, ranging from integrated renewable~~ concerns over utility bills, grid instability, and ~~storage systems to Microsoft’s plan to expand capacity~~ water usage—with single facilities consuming up to ~~38 gigawatts by 2032, and the use~~ 5 million gallons of ~~on-site power alternatives from providers like Bloom Energy~~ water daily. In response to ~~bypass grid delays.~~ local moratoriums, industry leaders have advocated for energy transparency standards to report efficiency metrics. The expansion is also reshaping financial and risk landscapes. The surge in AI-generated ~~data, including synthetic~~ data ~~and model logs,~~ has led 94.7% of surveyed organizations to increase data ~~storage, with 75% reactivating archives to feed models.~~ storage. This infrastructure boom is creating a massive market for the insurance industry; the Swiss Re Institute estimates that AI data centers and renewable energy projects could generate approximately $200 billion in insurance premiums between 2026 and ~~2030, with data centers alone accounting for $91 billion.~~ 2030. However, physical risks remain high, particularly in the ~~United States,~~ U.S., where facilities face exposure to severe weather like tornadoes and hail, with replacement costs for a single campus potentially reaching $50 billion.

Versions

Only revisions since CLSTR began indexing content versions appear here. Select a version to see what changed compared to the one before it.