Overview
- The buildout for AI computing is unusually large and costly with economists estimating trillions in spending and annual infrastructure outlays equal to a meaningful share of GDP over the coming years.
- State and local officials and organized community opposition are actively delaying, repurposing or blocking projects by using moratoriums, zoning fights and public protest, shifting the immediate power over siting from Washington to towns and counties.
- Rapid growth is raising concrete environmental and grid concerns because data centers use large amounts of electricity and cooling water, with projections showing a double‑digit share of U.S. power demand by 2030 in some analyses.
- Financing has moved from company cash to bonds, private credit and joint ventures, concentrating risk in outside investors and lenders and prompting investor caution as project debt and asset values face uncertainty.
- Policymakers are responding with measures that would make developers fund incremental grid work and disclose impacts, including a broadly supported House bill and a bipartisan framework urging developer‑paid upgrades and greater transparency.