Lee Zeldin, administrator of the Environmental Protection Agency (EPA), on Sunday warned against banning data centers in the United States, arguing that such restrictions could hamper American competitiveness in the artificial intelligence race against China.
Zeldin's comments came as several states have begun exploring moratoriums or bans on new data center construction, citing concerns about energy consumption, water usage, and environmental impacts. Data centers, which power cloud computing and AI systems, require significant electricity and cooling resources.
"There's been a push from some states where they're talking about a statewide moratorium," Zeldin said in an interview broadcast Sunday. "What I've found is that most of this country wants to make sure that we're winning the race on AI against China."
The EPA administrator emphasized national security implications, suggesting that restricting domestic data center development could cede ground to Chinese competitors in emerging technologies.
What the Right Is Saying
Conservative and business leaders largely support Zeldin's position, arguing that restricting domestic AI infrastructure would undermine American economic and national security interests. They contend that China has invested heavily in technology dominance and that regulatory barriers could hand Beijing a strategic advantage.
"Data centers are the factories of the 21st century," said one industry commentator who focuses on tech policy. "If we restrict our ability to build them here, we'll simply end up more dependent on foreign infrastructure."
Supporters of expanded data center development argue that concerns about energy use can be addressed through innovation rather than prohibition, pointing to advances in cooling technology and renewable energy adoption by major tech companies.
What the Left Is Saying
Environmental advocates and progressive policymakers who support data center moratoriums argue that the technology sector's expansion carries real costs for local communities and ecosystems. They point to increased electricity demand straining power grids, water consumption for cooling systems, and land use concerns.
Some Democratic legislators have called for stricter regulations on data center construction rather than outright bans, including requirements for renewable energy usage and efficiency standards. Environmental groups argue that the tech industry should bear the full cost of its infrastructure footprint.
"We need to be thoughtful about how we grow this infrastructure," said one environmental advocate who works on digital sustainability issues. "The question isn't whether we'll have data centers—it's whether they'll be built responsibly."
What the Numbers Show
Data centers consumed approximately 260 terawatt-hours of electricity in 2022, according to the International Energy Agency, representing about 1 percent of global electricity demand. That figure is projected to double by 2026 as AI applications expand.
The United States hosts the largest number of data centers globally, with major concentrations in Virginia, Texas, California, and Oregon. China ranks second worldwide in data center capacity, according to industry analyses.
AI workloads are estimated to require three to five times more computing power than traditional cloud services, driving accelerated demand for new facilities and infrastructure upgrades.
The Bottom Line
The debate over data centers reflects broader tensions between environmental regulation and technology competitiveness—a dynamic that has intensified as AI capabilities have become a focal point of U.S.-China strategic competition. Zeldin's comments signal the Trump administration's interest in preempting state-level restrictions on data center construction, though federal authority to override such bans remains legally unclear.
What happens next: States considering moratoriums include Virginia, where legislators have debated restrictions near sensitive areas, and California, where energy demand projections have prompted infrastructure reviews. The outcome of these state-level deliberations could shape the trajectory of domestic AI development for years to come.