On December 18, 2025, the Federal Energy Regulatory Commission (FERC) issued one of the most consequential orders of the year for grid operators, generation developers, hyperscalers, and large industrial customers. The order in Docket No. EL25-49 directs PJM Interconnection to rewrite how its tariff treats large loads, most notably AI-driven data centers, that co-locate with generating facilities behind the point of interconnection. Four months later, on April 16, 2026, FERC accepted in part and rejected in part PJM’s compliance filing and ordered yet another round of revisions.
The docket now spans more than 300 filings, multiple stakeholder coalitions, and a compliance phase. For energy professionals tracking resource adequacy, data center demand, and federal-state jurisdiction, that volume is a huge undertaking. HData’s Regulatory AI shows how energy professionals could handle such volume in minutes rather than weeks. That payoff starts with understanding what’s actually in the docket.
The case started with a complaint from Constellation Energy that PJM’s tariff was unjust and unreasonable. FERC took the concern seriously, issuing a February 20, 2025 show cause order challenging whether PJM’s Open Access Transmission Tariff (OATT)—the rulebook for moving power across PJM’s grid—fairly addressed co-located load. PJM and the Indicated PJM Transmission Owners pushed back on March 24, 2025, arguing the existing tariff was fine.
FERC disagreed and found PJM’s tariff unjust and unreasonable on two grounds: it lacks clarity on the rates, terms, and conditions governing interconnection customers that serve co-located load, and it offers no transmission service that reflects co-located loads’ ability to limit their withdrawals from the grid. Likewise, FERC found PJM’s behind-the-meter generation rules unjust and unreasonable for today’s large loads, citing cost-shift and reliability concerns.
FERC directed PJM to file tariff amendments by February 16, 2026, establishing three new transmission service options alongside Network Integration Transmission Service on a “gross demand basis” (billed on total load, not net of behind-the-meter generation) :
FERC also opened a paper hearing to determine just and reasonable rates, with briefs filed through mid-April 2026.
PJM, the Transmission Owners, and many intervenors agreed on the fundamentals: FERC has jurisdiction over transmission and ancillary service cost allocation with customers using or benefiting from the grid. Both FERC and PJM acknowledged that co-located loads must stay synchronized to the PJM grid and depend on it to operate, and that co-location raises legitimate concerns about reliability, study processes, and resource adequacy.
Where they disagreed was on the existing framework: treating co-located large loads as Network Load served through existing transmission products. PJM argued that is sufficient. FERC disagreed and ordered new service constructs the company had not proposed.
One issue FERC explicitly left open: broader jurisdictional questions about how retail loads served through co-location interconnect to the interstate system. Those questions will likely be addressed in the forthcoming Advance Notice of Proposed Rulemaking (ANOPR) on large load interconnection (RM26-4).
With more than 300 filings, four stakeholder groups, and a compliance phase on the move, EL25-49 is the kind of proceeding that is burdensome to track manually. We asked HData’s Regulatory AI a single question: “Based on where the docket is, what the intervenors have said, and how FERC has reacted, how are parties responding to the order?” Within two minutes, we received a structured response with citations linking back to each underlying filing.
Three stakeholder coalitions emerged from the response:
Generators and energy developers. Vistra Corp.’s response brief identified three flaws in PJM’s compliance proposals: that PJM’s proposed delay until 2029 effectively creates a three-to-five-year moratorium on co-location, undermining the “speed to power” needed for AI infrastructure; that charging NITS zonal rates for Non-Firm Contract Demand Transmission Service is discriminatory because the service is not superior enough to justify the higher cost over existing non-firm point-to-point service; and that PJM and the Transmission Owners reserved “unfettered discretion” over whether to conduct studies and on what timeline, violating the rule of reason.
Indicated PJM Transmission Owners. The Transmission Owners requested rehearing of the Co-Location Order, raising challenges on evidentiary basis, cost responsibility, the grid reliance charge, reliability risks, jurisdiction and operational penalties.
Public interest organizations. Earthjustice, the Environmental Defense Fund, and Sierra Club focused on PJM’s planning and load forecasting deficiencies, specifically forecast accuracy and the gap between projected costs and the customers actually paying them.
Regulatory AI also surfaced PJM’s Independent Market Monitor (IMM) as a distinct, non-party voice in the proceeding. The IMM strongly opposed Constellation’s request to avoid delisting capacity in PJM’s capacity auction, arguing that allowing a generator to sell the same capacity to both a co-located data center and PJM’s auction would let capacity be “sold twice,” increasing power costs for PJM customers and reducing reliability.
On FERC’s side, Regulatory AI identified FERC’s procedural moves with citations: the paper hearing scope, extension of time, denial of rehearing by operation of law, jurisdictional clarifications, and forward directives. It also flagged what FERC has not yet ruled on: the final Grid Reliance Charge percentage, the resolution of operational penalty disputes, and the ultimate implementation timeline for the new transmission services.
Several catalysts will shape the next phase of EL25-49:
As each of these develops, HData users can track every filing, order, and shift in position across the EL25-49 docket and the broader large-load proceedings now underway. To learn more or stay abreast of this docket with Regulatory AI, contact HData.
As the AI-native operating system for energy regulation, HData serves the largest customer ecosystem in regulated energy, helping utilities, regulators, advocates, advisory firms, corporates, and energy technology companies navigate regulatory complexity. Through centralized data, domain AI, and purpose-built applications, HData accelerates the research, analysis, and workflows critical to how the future of energy is decided.