bytevyte
bytevyte
Language
EN DE ES FR PT IT
Follow
ai-beats —

Data Center Electricity Tax Gains Ground as North Carolina Repeals Power Break for AI Facilities

data center electricity tax

North Carolina has repealed the sales and use tax exemption on electricity purchased by qualifying data centers, making it the second state after Virginia to directly impose a data center electricity tax. Governor Josh Stein signed the change into law this week as part of the state's 2026 budget. The new law keeps tax exemptions on hardware and other capital investments in place, but removes the electricity purchase break that had been one of the most generous operating-cost reductions for hyperscale facilities.

This shift is not an isolated event. It is the visible symptom of a policy inflection point that has been building across state legislatures since the AI buildout began straining power grids in earnest. The tax breaks that once seemed like reasonable economic-development tools now read differently when a single AI data center can draw as much electricity as a mid-sized town. Residents in multiple states have started asking who is paying for the grid upgrades those facilities require, and lawmakers are listening.

The Numbers Behind the Data Center Electricity Tax

The immediate fiscal impact is modest but telling. North Carolina's budget office estimates the repeal will generate $21.4 million in additional General Fund revenue in fiscal year 2026-27, rising to $28.6 million annually by fiscal year 2030-31. Those figures are small relative to the total state budget, but they signal a principle. Lawmakers are no longer willing to treat data center electricity as a tax-free input when the grid costs it imposes are borne by residential and small-business ratepayers.

Virginia set the precedent earlier this year with a first-of-its-kind consumption tax of $0.011 per kilowatt-hour on data center electricity. The rate is low enough that it will not deter new construction in Northern Virginia's Data Center Alley, but it establishes a framework other states can adopt. According to the National Conference of State Legislatures, at least 28 states introduced proposals during the 2026 legislative sessions to substantially amend their data center tax breaks, and at least nine considered full repeal.

The direction of the proposals, rather than their number, is what stands out. For more than a decade, states competed to offer the most generous data center incentives, treating every new facility as an unqualified economic win. Mississippi, Ohio, and Arizona all expanded tax breaks during the 2010s to attract server farms. That calculus has shifted. Lawmakers in 2026 are adding guardrails, including performance requirements, job-creation thresholds, and energy-efficiency mandates, that would have been unthinkable in the pre-AI era of data center recruitment.

The Broader Backlash Against AI Infrastructure

The movement extends well beyond electricity taxation. At least nine states have considered fully repealing their data center incentive programs, going beyond mere amendments to existing breaks. The pushback reflects a growing recognition that AI infrastructure carries hidden costs that the original incentive frameworks never accounted for.

Consider the ratepayer burden. When a hyperscale data center connects to the grid, the local utility must often upgrade substations, build new transmission lines, and maintain reserve capacity for what is essentially a 24/7 load. Under the old incentive model, the data center's owners paid no tax on the electricity itself, meaning the grid costs were socialized across all ratepayers while the benefits accrued to a single corporate tenant. That arrangement was sustainable when data centers were small and few. It is not sustainable when a single campus draws 200 megawatts or more.

Virginia's Northern Virginia region, home to the world's largest concentration of data centers, has become the laboratory for this tension. Dominion Energy has warned that data center demand could push the region's peak load past what its current infrastructure can support, triggering rate increases for residential customers. Maryland and Georgia have seen similar dynamics emerge as new campuses come online.

Why the Industry Objection Falls Short

The standard industry pushback to the data center electricity tax argument is that it will chase investment to other states or countries, slowing AI development and ceding competitive advantage. That objection overstates the risk for two reasons.

First, the scale of AI investment is now so large that a modest electricity tax is a rounding error in total project cost. A hyperscaler spending $5 billion on a single campus is not going to walk away over a tax worth a few million dollars per year, especially when power costs are already the largest operating expense. Second, the trend is national. With 28 states moving in the same direction, there is no tax-haven jurisdiction left to flee to. The competitive race has changed from who can offer the most generous subsidy to who can offer the most reliable grid.

The more serious risk is the opposite one: that states move too aggressively and create a regulatory patchwork that penalizes new construction before the grid upgrades are even complete. A careful, phased approach, such as North Carolina's repeal of only the power exemption while keeping capital incentives or Virginia's low per-kilowatt-hour consumption tax, strikes a more durable balance than a wholesale revocation of every incentive.

What Comes Next

The data center industry should expect more states to follow Virginia and North Carolina in 2027. The legislative trend is unmistakable, and the public pressure is not going to ease as long as residential electricity rates rise while hyperscalers receive tax exemptions on their largest operating cost. For utilities, the challenge will be to build out grid capacity fast enough to meet demand without saddling retail customers with the entire cost.

For the hyperscalers themselves, the calculus is changing. The era of cheap, subsidized power for AI training is drawing to a close. The cost of compute has always come down over time; the cost of the electricity that powers it is now heading in the opposite direction, at least in the jurisdictions where data centers concentrate. That structural shift will affect site-selection decisions and the overall economics of the AI industry.

Why This Matters

The North Carolina repeal is not a quirk. It is the leading edge of a broader recalibration in how states value AI infrastructure against the costs it imposes on ratepayers and grids. For hyperscalers and the utilities that serve them, the message is clear: the era of subsidized AI power is ending, and the costs of grid infrastructure will increasingly be priced into the buildout. That means higher operating costs for data centers, yes, but also a more honest accounting of what AI infrastructure actually costs the communities that host it. The data center electricity tax is becoming the new normal, and the industry should plan accordingly.

Photo by Winston Chen on Unsplash

✔Human Verified

Researched and cross-referenced against primary sources by the Bytevyte editorial team.

Share