Electric grid capacity auctions and ratings analysis are putting sharper numbers on the cost of the AI data center buildout, with rate impacts increasingly tied to how new large-load customers interact with grid planning rules. A report on PJM’s capacity auction outcomes described the 2028–29 clearing price hitting its maximum allowed cap even as the region faced a shortfall of supply. Moody’s Ratings flagged that grid mechanisms can socialize new generation build costs across all customers rather than charging them directly to the new AI entrants. The reporting notes that data centers are a primary driver of capacity charges, implying households and small businesses could absorb more of the infrastructure bill. For higher-education leaders, the takeaway is operational: campus energy procurement, demand management, and sustainability planning increasingly intersect with external grid pricing dynamics driven by AI infrastructure growth.