by Kelly KIRSCH Directeur Général
Introduction: The Hidden Infrastructure Behind AI
Artificial intelligence may seem abstract, but its foundation is anything but. Across the United States, data centers—massive warehouse-sized facilities filled with servers—are rapidly multiplying to support the explosive growth of AI.
More than 4,000 data centers are already operational, primarily in states like Virginia, Texas, and California, with another 3,000 planned or under construction. This surge is driven by skyrocketing demand for AI and incentivized by generous state tax breaks.
For tech companies and policymakers, this expansion is essential to maintaining America’s position as a global AI leader. But for many communities, the reality looks very different.
The Promise: Jobs, Growth, and Technological Leadership
Data center developers and tech giants present a clear narrative: these projects bring economic opportunity.
They argue that data centers:
In a geopolitical race against China and other global powers, expanding digital infrastructure is framed as not just beneficial—but necessary.
The Reality: A Growing Backlash
Despite these promises, public opposition is rising across the country—and for good reason.
Communities are increasingly concerned about:
Experts like Ben Green, a professor at the University of Michigan, argue that these concerns are not exaggerated—they are well-founded. He describes the situation as a “David and Goliath” battle, with local communities pushing back against some of the world’s most powerful tech companies.
The Resource Strain: Energy and Water at Scale
Modern data centers are no longer modest facilities. They are enormous.
A single hyperscale project in Michigan—part of OpenAI’s Stargate initiative—is expected to span over 2 million square feet and consume 1.4 gigawatts of energy, roughly equivalent to powering one million households.
Looking ahead:
This rapid expansion is not just about individual facilities—it’s about the cumulative impact of thousands of them nationwide.
The Jobs Myth: Short-Term Gains, Long-Term Limits
One of the most compelling arguments for data centers—job creation—does not fully hold up under scrutiny.
While construction projects generate temporary employment, the operational phase requires surprisingly few workers:
Unlike tech campuses, data centers are not hubs of innovation or large-scale employment—they are automated infrastructure.
This gap between promise and reality has fueled skepticism among both policymakers and residents.
The Economic Trade-Off: Tax Breaks vs. Public Benefit
States have aggressively courted data center investments with tax incentives. Today, around 35 states offer tax breaks to attract developers.
But the cost is significant:
Even the economic benefits that do materialize often fall short of expectations, with limited spillover into local economies.
A Bipartisan Issue: Uniting Unlikely Allies
Opposition to data centers is notable for one reason—it crosses political lines.
This rare alignment has turned data centers into a key electoral issue, influencing local and state elections in places like Virginia, Georgia, and Michigan.
Regulation on the Rise: A Turning Point
As public pressure grows, governments are beginning to act.
In a landmark move, Maine lawmakers approved the nation’s first statewide moratorium on new data centers, pausing approvals for 18 months to study their impact.
Other states are considering similar restrictions, focusing on:
However, this push for regulation faces resistance at the federal level, where policymakers warn that overregulation could undermine America’s competitive edge in AI.
The Policy Challenge: Balancing Innovation and Impact
Experts suggest several key reforms to address the growing tensions:
1. End or Reduce Tax Incentives
Eliminating excessive tax breaks could slow unchecked expansion and ensure fairer economic outcomes.
2. Increase Transparency
Contracts and agreements should be public, with fewer redactions and clearer accountability.
3. Protect Consumers
Utility upgrades should not lead to higher electricity bills for residents.
4. Empower Local Communities
Residents should have a stronger voice in whether—and how—data centers are built.
5. Plan for Sustainability
Governments must assess how many facilities their infrastructure and environment can realistically support.
Conclusion: The Cost of Powering the AI Future
There is little doubt that AI will continue to expand—and with it, the need for data centers.
But the current trajectory raises a fundamental question:
Who benefits—and who pays the price?
For tech companies, data centers are the engines of innovation.
For communities, they can mean higher costs, environmental strain, and limited economic return.
The challenge ahead is not whether to build the infrastructure for AI—but how to do so responsibly.
Because in the race for technological supremacy, the real test may not be how fast we build—but how wisely.