Virginia is where the US data-center boom is easiest to see in the electricity statistics. The US Energy Information Administration reported in May 2026 that commercial electricity sales in the state had risen by nearly 30 million megawatt-hours between 2019 and 2025, with data centers responsible for much of the increase.
That growth is extraordinary because it is no longer a niche industrial load. Data centers are large enough to alter the shape of statewide commercial electricity demand, influence utility investment plans and change the politics of new transmission and generation.
Northern Virginia's advantage was built before the AI boom
The Northern Virginia cluster grew around dense fibre connectivity, proximity to major internet exchange points, cloud customers and a long history of large-scale data-center development. Those network effects are difficult for a new market to reproduce quickly.
AI has intensified rather than created the cluster's infrastructure challenge. High-density accelerator deployments can require far more power per unit of floor space than conventional enterprise computing, making electrical capacity increasingly more important than building square footage.
Electricity demand changes the economics for everyone
When a small number of campuses add gigawatt-scale demand, utilities need generation, transmission, substations and long-lead equipment. The cost and timing of that infrastructure become public-policy issues because households and other businesses also depend on the same system.
The relevant question is therefore not whether data centers are economically valuable in isolation. It is how grid costs are allocated, whether new large loads bring enough generation or infrastructure with them, and whether permitting can keep pace without shifting disproportionate costs to other ratepayers.
Virginia's tax model is part of the debate
Virginia has historically used sales-and-use tax exemptions for qualifying data-center equipment as an investment incentive. As the sector has grown, lawmakers and policy analysts have increasingly debated whether incentive design should change to reflect power demand, job creation and local infrastructure costs.
A mature cluster can justify different policy questions from an emerging one. Once companies locate in Virginia because of network density and existing ecosystems rather than only tax treatment, the state has more room to ask how new projects contribute to the infrastructure they consume.
Why Virginia matters to the national AI economy
Virginia is a useful benchmark for every other state pursuing AI infrastructure. It shows what happens when data centers become a structural electricity load rather than a collection of individual construction projects.
Texas may have a larger greenfield investment pipeline and California a deeper AI-company base, but Virginia provides the clearest operating evidence of compute reshaping the power system. ACR will use this page as the demand-side counterpart to its national data-center investment tracker.
| Indicator | Evidence | Interpretation |
|---|---|---|
| Commercial electricity sales, 2019-2025 | Increase of nearly 30 million MWh | Large structural load growth |
| Primary growth driver | Data centers, according to EIA | Compute now affects statewide electricity planning |
| Core cluster | Northern Virginia | Deep fibre, cloud and enterprise ecosystem |
| Main constraint | Power and transmission | Capacity increasingly determines growth |
| Policy lever | Data-center tax incentives | Growing debate over costs and benefits |
Frequently asked questions
How much has electricity demand grown in Virginia because of data centers?
EIA reported that Virginia commercial electricity sales rose by nearly 30 million megawatt-hours from 2019 to 2025, with data-center growth responsible for much of that increase.
Why is Northern Virginia such a large data-center market?
The region combines dense fibre networks, internet connectivity, a long-established cloud ecosystem and a large base of enterprise and government demand. Those network effects accumulated over many years.