Virginia Governor Issues EO on Data Center Accountability Framework

by Zachary Barlow

September 22, 2026

Last week, the Governor of Virginia signed Executive Order 22 (EO 22). EO 22 establishes new constraints and regulations for data center developments in response to constituent demands. A recent Williams Mullen memo discusses the key requirements under the new data center accountability framework:

“EO 22 is the most sweeping executive action on data center development in Virginia to date, imposing immediate restrictions on state economic development assistance, banning NDAs related to data center projects, launching environmental and community-impact initiatives, and creating a rapid-response AI Task Force. While many EO 22 provisions take effect immediately, the Governor has announced she will work with the General Assembly in the 2027 legislative session to codify these standards into permanent law and pursue additional legislative guardrails.”

While this data center accountability framework doesn’t go as far as others that we’ve seen, it still imposes significant restrictions on developers. Stopping short of denying permitting altogether, this EO does restrict data centers’ access to accelerated permitting programs. Additionally, it attempts to address water and electricity consumption issues by creating new rules for water-scarce areas and favors permitting for projects with on-site energy generation. In addition to establishing this framework in the immediate term, the EO also looks to leverage the legislature in Virginia’s next legislative session.

Data centers are growing increasingly unpopular, and constituents and politicians alike are asking more questions about how development will impact the surrounding areas. Data center developers must keep environmental impacts at the top of mind and be willing to make concessions in the community’s interests. The ongoing controversy surrounding data center development threatens AI companies’ social license to operate. This could lead to bottlenecked compute power and raise the price of using AI. Companies should be careful not to overrely on large amounts of cheaply available compute. If AI is integrated into your core functions, then the rising costs of AI could present serious risks.