The Data Center Boom Is Stalling: Why Resistance Is Working

There's a narrative being sold that AI data centers are popping up everywhere, unstoppable and inevitable. The reality, as CNN's David Goldman reports, is far more complicated — and far more encouraging for communities fighting back.

Yes, the planned boom is massive. But surprisingly few of these data centers are actually getting built, thanks to a combination of public opposition, permitting hurdles, and supply chain bottlenecks that the industry can't overcome.

The Numbers Tell a Different Story

The American public has made its position clear. A Gallup poll showed 71% of Americans oppose AI data centers in their area. Politicians are running against them, and several counties and states have proposed or passed laws to ban them.

But opposition is only part of the picture. The industry's own numbers reveal a pipeline that's buckling under its own weight:

• Only about half of the AI computing capacity scheduled to activate between now and 2028 is expected to come online on time, according to Goldman Sachs.
• Despite $750 billion in AI infrastructure investments this year alone (JPMorgan), about 60% of data center capacity planned for 2027 hasn't even begun construction.
• The US had 5,427 data centers at the end of last year (Stanford AI Index Report). AI companies have announced plans for 3,969 new ones — but just 802 are currently under construction (Aterio).
• Of the 565 gigawatts of computing power AI companies are planning, Columbia Business School professor Stijn Van Nieuwerburgh expects just 180 gigawatts to actually get built over the next decade. Two-thirds of the pipeline is "implausible."

What's Slowing the Boom

Materials shortages: Building materials are difficult to source, and the chips these facilities house are in short supply. Stanford's AI Index report calls Taiwan's TSMC — which fabricates virtually every leading AI chip — "a single point of dependency in the global AI supply chain."

Power shortages: Data centers already account for roughly 8% of US electricity usage, and that could grow to 12% by 2028. Wait times for generation step-up transformers have tripled. GE Vernova, the largest natural gas turbine manufacturer, reported that bookings for its power generators have doubled to $200 billion over five years.

Labor shortages: Meeting the proposed buildout deadlines would require adding 500,000 electricians, 300,000 welders, and 550,000 plumbers — workers who simply don't exist in the pipeline.

Public opposition: About a dozen states have proposed data center building moratoriums. New York and Texas recently enacted temporary bans. But the biggest hurdle isn't even the bans — it's getting construction permits approved, which Goldman Sachs identifies as the primary bottleneck.

The Takeaway

Even with $10 trillion of projected investment — 50% bigger than the 19th century railroad expansion — the AI data center boom is hitting a wall. The industry's own projections are collapsing under the weight of physical reality: there aren't enough materials, enough power, enough workers, or enough community consent to build what they've promised.

That last factor matters most. When communities organize, when states pass moratoriums, when permitting processes are scrutinized — the pipeline shrinks. The narrative of inevitability is a tool of the industry. The data shows that resistance works.

Source: CNN — "Americans are rallying against data centers. Surprisingly few are actually getting built" by David Goldman, August 6, 2026.