AI Governance
Data Centers
Drones
Climate
Privacy
2026 Maine Senate
Bob Dylan
Who Pays for AI’s Power Demand?
Coverage from AOL, Institute of Energy and the Environment, and others
Rapid AI data center expansion is increasing projected electricity demand and putting pressure on generation, transmission, capacity markets, and grid reliability across major U.
S. regions. In PJM, capacity prices and estimated ratepayer costs have risen sharply, while utilities and regulators debate how much of the expense should be assigned to large data center customers rather than households and manufacturers. At the same time, interconnection delays, uncertain project pipelines, governance disputes, and local opposition are slowing or challenging planned developments.
The story broadens from quantified PJM cost pressures to greater uncertainty around interconnection, construction timing, siting, and the generation technologies needed to serve new loads. Federal involvement is also more explicit through the White House’s pressure on PJM.
The story now includes a concrete statewide regulatory response: New York temporarily halted hyperscale data center permitting. It also sharpens the supply-side constraint narrative, emphasizing that gas-generation plans face turbine shortages, water limits, costs, and permitting delays.
The story has shifted from primarily an interconnection and construction bottleneck to a broader electricity-cost and reliability crisis, with regulators and communities taking more active steps to constrain projects. A NERC alert and sharper scrutiny of speculative pipeline estimates add urgency to concerns about whether projected capacity will materialize safely and affordably.
- A NERC alert followed rapid data center load disconnections that raised grid reliability concerns.
- Households and manufacturers in Ohio and Arizona joined Pennsylvania in prominent cost-allocation disputes.
- States and communities are implementing audits, moratoria, and enhanced oversight of proposed data centers.
- Analysts now identify speculative requests, duplicate counting, and inexperienced developers as reasons pipeline estimates may be overstated.
- Proposals increasingly require data centers to cover exit costs alongside generation, transmission, and capacity costs.
AI data center projects are expanding faster than electricity supply, grid connections, and local approval processes can accommodate. Developers are pursuing onsite generation and faster transmission access, while regulators and utilities debate who should pay for upgrades and communities challenge projects over electricity prices, noise, pollution, water use, and land impacts. The result is greater uncertainty around project timelines, operating reliability, and whether planned capacity will be built at the announced scale.