AI Power Grab Fails to Deliver: Phantom Projects Clog US Grid
The US artificial intelligence industry's insatiable appetite for power has led to a staggering 1,066 gigawatts of requested electricity capacity to support data center growth. However, a significant portion of these requests, approximately 72%, are considered 'phantom' projects, with no realistic path to construction.
Developers submit applications to multiple utilities simultaneously, shopping for the best grid connection terms, and each utility logs the full requested load as a real demand signal, unaware that others are holding identical applications from the same developer. This results in a queue stuffed with mirages, making it difficult for utilities and grid operators to distinguish genuine demand from noise.
Utilities have started responding by shifting to 'first-ready, first-served' frameworks that require meaningful upfront financial commitments before an application is treated as a real demand signal. Higher fees, credit requirements, and collateral deposits are becoming standard filters designed to separate serious developers from queue-stuffers.