New York’s new data center moratorium misunderstands artificial intelligence.
New York’s new data center moratorium misunderstands artificial intelligence.
L ast week, Anthropic researcher Jacob Coxon resigned, warning of an “endgame” in which artificial intelligence companies could help determine humanity’s fate. It is the latest apocalyptic warning to fuel fear and misunderstanding around AI, but it matters because speculative fears can produce very real policy mistakes. New York offers a timely example: Governor Kathy Hochul recently imposed the nation’s first statewide moratorium on new hyperscale data centers — the infrastructure that powers AI and much of the modern digital economy.
At the very moment artificial intelligence is emerging as one of the defining innovations of our time, New York is choosing to halt the infrastructure that makes it possible.
Proponents of the moratorium defend it as a temporary review of energy and environmental impacts — and those issues deserve attention. Communities should have a voice in how major projects affect local electricity costs, water resources, and quality of life. That is not “Not in My Backyard” (NIMBYism); it’s accountability. Businesses and elected officials have an obligation to address legitimate community and customer concerns; that’s how markets and democracy work. But a blanket statewide moratorium is different. Halting data center construction based on bad information or irrational fear is not careful planning; it is a harmful pause on the infrastructure of innovation and the economic future.
That pause reflects a broader problem in the AI debate. New technologies are by nature disruptive, and it is reasonable for people to ask how AI will affect their lives — policymakers should not ignore this. But policy should be based on an understanding of what AI actually does, the infrastructure it requires, and its concrete benefits and risks.
AI is a tool that analyzes patterns from enormous amounts of data to predict the next word, phrase, image element, or line of code in a user’s request. That predictive ability is what allows it to draft, translate, and write code so quickly. It is powerful not because it possesses human judgment, but because it can process information at a speed and scale no individual human can match.
AI’s most immediate promise is not replacing human beings but augmenting human work — helping small businesses reduce administrative burdens, doctors’ offices organize records, nonprofits analyze information, and governments deliver services more efficiently and accurately. Every one of those tasks is handled by powerful computers housed in data centers, supported by servers, cooling systems, fiber connections, and electricity. The potential usefulness of AI cannot be achieved without this infrastructure.
Too many policymakers want the promise of AI while resisting the data centers and energy systems that have made our modern life and the rise of AI tools possible. This is the technological equivalent of wanting air travel without airports or online commerce without warehouses.
Data centers are not new. They already support banking, streaming services, health records, logistics, cloud storage, cybersecurity, and countless other parts of modern life. AI did not create the need for data centers; our world already depends on them.
When responsibly built, data centers can deliver meaningful local benefits, including construction jobs, permanent technical and facility jobs, and substantial property-tax revenue that can fund schools, public safety, infrastructure, and other community priorities. It is true that like every other piece of modern productive infrastructure, data centers require electricity to operate. But if New York faces limited energy supply, the answer is faster permitting for energy projects and a stronger grid, not blocking creation of the infrastructure of the next economy.
New York should also remember its own recent history with novel technology. In 2014, the state banned hydraulic fracturing, the drilling technique that helped unlock natural gas from the Marcellus Shale formation. That ban pleased activists, but communities in New York’s Southern Tier watched neighboring Pennsylvania benefit from energy investment, jobs, and tax revenue that New York chose to forgo. Data centers are not natural-gas wells, but the lesson is similar: blanket bans rarely eliminate economic opportunity. They usually move it somewhere else.
New York’s own history of building, investment, and innovation made it a global center of commerce. Previous generations understood that prosperity requires investment and infrastructure. Modern energy systems, digital infrastructure like data centers, and sensible permitting are the foundation on which America’s future prosperity will be built.
This does not mean “no guardrails or rules.” When AI is used to commit fraud, theft, discrimination, or privacy violations, the first question should be how existing law applies — not how quickly lawmakers can create another blanket restriction, regulation, or law. That said, reasonable community concerns about data centers should be addressed.
What policymakers should not do is let 50 states create conflicting rules or bans before the technology and its uses are fully understood. Premature, fragmented regulation built on misunderstanding and fear locks in false assumptions, raises costs, and makes it harder for workers, businesses, public agencies, and consumers to benefit from technology.
New York’s data center moratorium is not just a decision about environmental or land-use impacts. It is a test of whether policymakers will respond to technological change with understanding or fear — and whether New York will build the infrastructure of the next economy or send investment, innovation, and opportunity elsewhere.