First-order effects

• Utilities seeking approval can begin recovering proposed grid-investment costs through customer rates if regulators approve them; households and businesses in affected territories face higher-bill exposure.
• Data-center-driven load growth becomes a formal justification in rate proceedings, putting utilities, regulators, and large new power users into the same cost-allocation debate.

Second-order effects

• Projected bill increases in PJM territory heighten pressure on regulators to scrutinize whether new large loads should bear more of the incremental grid cost rather than spreading it across existing customers.
• Higher and less predictable power costs can become a siting and operating constraint for data-center developers, alongside the physical availability of grid connections.

Third-order effects

• If these cases persist, AI infrastructure buildouts will increasingly be governed by utility-rate design and interconnection rules, not only by chip, server, and data-center capacity.
• The enduring policy question is whether grid upgrades are treated as broadly socialized infrastructure or assigned more directly to large incremental users; the answer could reshape where compute capacity is built.