Public opinion in the United States has turned sharply against the construction of new data centers, with a recent survey indicating that only 15 percent of Americans would support such a facility in their area, compared to 75 percent in opposition. This marks a significant shift from the previous year when 43 percent approved and 42 percent opposed. This growing backlash is manifesting in political actions, with gubernatorial candidates in Wisconsin debating the issue and Pennsylvania Governor Josh Shapiro enacting new restrictions on AI infrastructure projects. Even Texas, known for its pro-business stance, has implemented a temporary moratorium on new data center approvals.
Data center projects are facing cancellation at a record rate due to mounting local opposition. However, despite this increased resistance, the expansion of AI infrastructure shows no signs of slowing. Industry experts and construction data suggest that the economic drivers behind the data center boom remain more powerful than the political opposition. Unless federal legislation imposes a nationwide moratorium, AI companies are expected to continue finding jurisdictions willing to host these facilities in exchange for revenue and incentives.
The Scale of the Buildout
The data center boom is characterized by two key facts: a significant increase in public opposition and an even larger scale of development. As of late July, over 500 jurisdictions have enacted severe constraints or outright bans on data center construction, with most of these measures implemented since the start of the year. The rate of cancellation for contested data center projects has also risen dramatically; in the first half of 2026, approximately 50 percent of disputed developments were canceled, compared to 20 percent in late 2024. By Heatmap's count, more than 100 projects have been canceled this year due to local opposition, with over 200 currently facing challenges.
Despite these numbers, these restrictions have not significantly impacted the overall AI buildout. More than 90 percent of U.S. counties still have no significant constraints on data center development. Furthermore, some recent policy changes, such as Texas's pause on approvals, do not halt construction for projects that can self-power using on-site natural gas. While states like Pennsylvania and New York have enacted more substantial restrictions, they represent a small fraction of the overall data center market. Combined, these two states host only about 6.5 percent of America's data centers.
In the first quarter of the year, at least 3.5 gigawatts of data center capacity were canceled due to local opposition. However, during the same period, at least 36 gigawatts were added to the U.S. pipeline of proposed and active projects. Industry analysts note that while cancellations have increased, the ratio of blocked to active developments has not fundamentally changed. Monthly construction spending on data centers in the U.S. reached a record high in June.
Driving Forces Behind Resilience
The resilience of the AI buildout can be attributed to several factors. Primarily, the demand for computing power is immense. As AI companies develop larger models and the use of artificial intelligence increases, the economy's need for computation has far outstripped supply. A recent report indicated that 99 percent of North America's data centers are occupied, with 95 percent of ongoing construction already reserved.
Secondly, data center projects are highly location-flexible. Unlike other industries that may face zoning restrictions, data centers can be built in remote areas. While some proximity to users is necessary for certain applications, hyperscale facilities used for training AI models can be located virtually anywhere with access to land, fiber optics, labor, and electricity. This flexibility, combined with high demand, allows companies to offer significant incentives to localities to secure project approvals. Many rural counties are also seeking revenue and investment, making them amenable to hosting data centers.
This dynamic suggests that without a federal moratorium, development deals are likely to continue. Investors note that companies may need to increase community benefits to gain acceptance. For example, West Virginia's governor has signaled plans to encourage data center development, using the revenue to reduce the state's income tax.
Broader Implications
The growing opposition raises questions about the aggregate impacts of data centers, such as increased carbon emissions and potential job displacement due to AI advancements. For climate activists, shifting development to areas with less renewable energy could be counterproductive. They may find that regulating data center development to ensure clean energy use and infrastructure expansion is more effective than outright bans.
For those concerned about the existential risks of AI, local bans may offer limited instrumental value. While restrictions can slow the buildout, the underlying economic forces and location flexibility mean that AI infrastructure will likely be built elsewhere. Ultimately, mitigating AI risks may depend more on regulations governing the use of AI and the distribution of generated wealth than on the physical location of data centers. The resistance movement is significant, but it faces the challenge of a massive investment frenzy, making a complete halt to the buildout unlikely without a collapse in demand or a major political shift.