US Senate Pushes for Mandatory Energy Reporting From Data Centers

Two US senators have formally requested the Energy Information Administration to establish mandatory annual energy reporting requirements for data centers. Senators Josh Hawley and Elizabeth Warren sent a letter to EIA administrator Tristan Abbey on Thursday, citing growing concerns about the impact of large load energy consumption on grid planning and oversight. The move signals a broader shift in how legislators are beginning to treat data center energy use, not as a peripheral environmental concern, but as a systemic risk to the reliability and planning integrity of the national grid.

The letter arrives against a backdrop of accelerating legislative activity targeting the data center industry. Earlier this week, Senator Bernie Sanders and Representative Alexandria Ocasio-Cortez announced plans to introduce legislation that would halt new data center construction until Congress reaches agreement on how to regulate AI. While that proposal targets construction directly, the Hawley-Warren letter takes a different approach, focusing on information collection as the foundation for any credible regulatory framework. The distinction matters because mandatory reporting, if implemented, would give federal planners visibility into energy consumption patterns that no standardized form currently captures anywhere in the federal data infrastructure.

What the Senate Wants From Data Centers

The senators are asking for granular, standardized data that the EIA does not currently collect. Their requests include hourly, annual, and peak energy loads, the rates companies pay utilities, details on grid upgrades triggered by new large loads, and whether data center operators participate in demand response programs. The specificity of the request suggests that the senators have consulted with grid planners and energy economists who understand where the current data gaps are creating planning blind spots for utilities and federal regulators alike.

Hawley and Warren also want the EIA to distinguish between energy consumption from AI computing tasks and general cloud services, a distinction that current federal reporting frameworks do not make. This differentiation is operationally significant because AI workloads generate fundamentally different demand profiles than conventional cloud computing, drawing sustained near-peak power for extended periods rather than the variable loads that characterize general enterprise compute. Without that distinction in reported data, grid planners cannot accurately model how AI infrastructure growth will affect electricity demand at the regional level over the next decade.

“As electricity demand growth continues to accelerate after years of relative stagnation, the lack of reliable, standardized data on large load energy consumption poses significant risks to effective grid planning and oversight,” the senators wrote in the letter, which TechCrunch reviewed.

Grid Pressure Behind the Push

The timing reflects real urgency. Energy use by data centers has grown sharply in recent years, with planned new facilities on track to nearly triple the sector’s total energy demand by 2035. The EIA currently tracks energy use across only four broad categories: residential, commercial, industrial, and transportation. Data centers fall under commercial with no separate classification, which means the federal government currently has no standardized mechanism for understanding how much electricity the sector consumes, how that consumption is distributed across markets, or how it is likely to grow as AI infrastructure investment accelerates.

The gap between what federal data infrastructure captures and what grid planners actually need to make sound investment decisions is becoming operationally consequential. Utilities making long-range transmission investment decisions are working with incomplete demand forecasts because the data center sector, which is now one of the fastest-growing sources of electricity demand in the country, does not appear separately from the broader commercial category. That forecasting gap translates directly into under-investment in transmission capacity in markets where AI infrastructure demand is concentrated, which in turn contributes to the interconnection queue backlogs that are slowing data center development across the country.

Abbey acknowledged in December that launching a new survey from scratch takes approximately two years through standard processes, though he noted that surveys of smaller scope can move faster under existing authorities. Hawley and Warren have requested a formal response from the EIA by April 9. Whether the agency moves forward with a mandatory reporting framework, and on what timeline, will depend partly on whether the current administration treats data center energy transparency as a grid reliability priority or a regulatory burden on an industry it has been broadly supportive of.