TellWell
…
← Misinformation tracker
UnverifiableOther · Science

Yes, U.S. Data Centers Are Booming to Meet the AI Surge — And the Numbers Are Staggering

“Data centers across the U.S. have emerged to meet the AI surge”

The argument in brief

The claim that data centers across the U.S. have emerged to meet the AI surge is true. Construction has hit record highs, with total U.S. primary market capacity exceeding 5,000 megawatts. Goldman Sachs projects a 165% increase in data center power demand by 2030, driven largely by AI workloads.

The numbersU.S. Primary Data Center Market Power Capacity (MW) - Key Markets

Data: CBRE North America Data Center Trends H2 2023

Why it spread

This claim spread easily because it fits a narrative people can already see around them — constant headlines about AI investment, tech giants announcing massive spending, and energy companies warning about grid strain. When a claim matches what feels intuitively obvious, people accept it quickly, sometimes without checking whether the scale and specifics hold up. In this case, they do.

The claim is true, and the scale of it is hard to overstate. U.S. data centers are being built at a record pace, and artificial intelligence — particularly the training and running of large language models — is the primary reason why.

CBRE Research tracked U.S. primary market data center inventory surpassing 5,000 megawatts in 2023, with construction pipelines at all-time highs. Key hubs like Northern Virginia, Dallas, Phoenix, and Chicago are all seeing major buildouts, according to JLL Research, driven directly by hyperscalers like Microsoft, Google, Amazon, and Meta pouring money into AI infrastructure.

The energy implications are enormous. The U.S. Department of Energy's Lawrence Berkeley National Laboratory identified AI workloads as a primary driver of new construction and rising electricity consumption nationwide. Goldman Sachs projects data center power demand will rise 165% by 2030, with hundreds of billions in new U.S. investment already planned. McKinsey estimates global data center investment could reach $5 trillion through 2030, with the U.S. taking a large share.

To be fair, not every new data center is purely AI-driven. Cloud computing, streaming, and general enterprise demand all play a role. But the evidence is clear that the AI surge has been the decisive accelerant behind the current construction boom — utilities and grid operators are already revising long-term capacity plans to keep up.

This story is worth watching closely because rapid data center expansion carries real consequences: surging energy demand, pressure on local power grids, water use for cooling, and land competition. As AI investment continues, communities near major data center corridors will feel the effects directly.

Sources

  • CBRE Research

    U.S. data center capacity under construction reached record highs in 2023, with primary markets seeing absorption surge driven by AI and cloud demand. Total U.S. primary market inventory exceeded 5,000 MW.

  • U.S. Department of Energy (Lawrence Berkeley National Laboratory)

    U.S. data center electricity consumption has grown substantially, with AI workloads identified as a primary driver of new construction and capacity expansion across the country.

  • Goldman Sachs Research

    Goldman Sachs projected a 165% increase in data center power demand by 2030, largely attributable to AI, with hundreds of billions of dollars in new data center investment planned across the U.S.

  • JLL Research

    JLL reported record data center construction pipelines in 2023-2024 across U.S. markets including Northern Virginia, Dallas, Phoenix, and Chicago, directly tied to hyperscaler AI infrastructure buildout.

  • McKinsey & Company

    McKinsey estimated that global data center demand could require $5 trillion in investment through 2030, with the U.S. accounting for a large share, driven primarily by generative AI adoption.

TellWell AI

Related debunks