The Federal Energy Regulatory Commission’s 60-day deadline expires on August 17, 2026, compelling six of the United States’ largest grid operators to either defend their current interconnection tariffs or propose sweeping changes to accelerate AI data center connections. This unprecedented regulatory action could reshape how large-scale AI infrastructure gets built in America for the next decade.
PJM Interconnection, MISO, SPP, CAISO, ISO-NE, and NYISO — the six regional operators controlling the US transmission grid — must file their responses to FERC’s show-cause orders by the deadline. The orders, issued unanimously on June 18, 2026, under Section 206 of the Federal Power Act, represent a rare use of binding regulatory authority rather than the typical multi-year rulemaking process.
The Power Crisis Behind the Orders
AI data centers consume extraordinary amounts of electricity. A single large-scale AI training facility can draw as much power as a small city, and the projected demand from hyperscalers’ expansion plans threatens to overwhelm grid infrastructure that took decades to develop. Connection queues across the country have grown to multi-year waits, with some projects facing 2029 or 2030 interconnection dates.
FERC’s action addresses a fundamental bottleneck: transmission interconnection processes designed for gradual load growth cannot accommodate the exponential demand from AI infrastructure. The commission’s show-cause mechanism forces operators to either demonstrate their current tariffs can handle rapid large-load growth or propose specific changes to accelerate approvals.
What’s at Stake
The operators’ responses will determine whether AI companies can build data centers where and when needed — or face years of delays that could shift competitive advantage to regions with more responsive grid governance. Key issues include:
- Queue management: First-ready, first-served vs. historical approaches that allow projects to hold positions without building
- Financial transmission rights: How to allocate costs among developers, utilities, and ratepayers
- Interconnection timelines: Whether processes can be compressed from years to months
- Capacity planning: How operators account for uncertain but potentially massive future demand
Industry Response
The AI industry has responded with aggressive infrastructure investments, including on-site power generation, nuclear PPAs, and proposals for gigawatt-scale campuses. But these investments require grid interconnection to deliver value — making the FERC deadline a critical inflection point for the sector’s growth trajectory.
Utility executives have expressed concern about the pace of demand, noting that transmission infrastructure cannot be built at the same speed as AI compute. Environmental advocates raise separate questions about the carbon footprint of unchecked data center growth, particularly in regions with fossil fuel-dependent grids.
What Happens Next
If operators demonstrate their current processes can handle AI data center demand, FERC may allow the status quo to continue. If not — which most industry observers expect — the commission will require specific tariff modifications, potentially accelerating interconnection timelines and shifting cost allocation formulas.
For AI companies planning multi-year infrastructure investments, the August 17 deadline offers the first concrete signal of whether the US regulatory system can support their ambitions — or whether they’ll need to pursue alternatives like international expansion or on-site generation to secure sufficient power for their compute needs.