Data centers are concentrated in these states. Here's what's happening to electricity prices (www.cnbc.com)

U.S. residential electricity bills are rising and data centers — especially large AI/ML facilities concentrated in a few states — are increasingly blamed. Nationwide bills were up about 6% in August year‑over‑year, but states with heavy data center footprints saw steeper jumps (Virginia +13%, Illinois +16%, Ohio +12%). Virginia hosts the world’s densest cluster of data centers and individual facilities can draw a gigawatt or more (roughly the load of 800,000 homes). Political pressure is mounting: state leaders and senators are publicly challenging tech companies over perceived subsidies and cost-shifting to consumers. The technical driver is grid capacity strain in PJM Interconnection, which serves many of those states. Capacity-auction costs exploded from $2.2B for 2024–25 to $14.7B for 2025–26 and rose to $16.1B in the latest round; Monitoring Analytics attributes roughly $9.3B (63%) of the 2025–26 bill to data center demand. Those capacity costs flow into consumer bills and have sparked regulatory scrutiny and potential policy changes that could affect siting, procurement, and economics of AI infrastructure. Regional contrasts matter: ERCOT (Texas) can connect supply faster (~3 years) and saw only ~4% price growth, while California’s high rates are driven by wildfire-prevention costs. For AI/ML operators, this means rising operating costs, greater permitting and grid-integration hurdles, and heightened political risk to large-scale expansion.