Kevin Hassett says US economy can thrive despite data center slowdown

Kevin Hassett wants Americans to know the economy doesn’t need every new server farm to get built. The White House adviser said the US economy can thrive even if data center growth slows.

What Hassett is arguing

Hassett, who directs the White House National Economic Council, has been one of the loudest voices tying US growth to private investment in AI. Data centers sit at the center of that pitch.

His own numbers show how big the bet is. Hassett has estimated that AI investment in the US could reach $10.3 trillion between 2025 and 2032.

That works out to an annual average of 3.63% of GDP. Put plainly, roughly three and a half cents of every dollar the economy produces each year would flow into AI buildout under that estimate.

Hassett has paired those projections with an upbeat growth forecast. He has held to a GDP growth outlook of around 4%, assuming no unexpected external shocks hit the economy.

He has also pushed back on fears that the AI trade is fragile. Despite recent swings in tech stocks, Hassett said the sector remains well-funded and resilient.

“No risk of a wholesale sell-off.”

The 1990s comparison

Hassett has compared the current AI boom to the internet explosion of the 1990s. In his framing, heavy spending on data centers, semiconductors and related technologies should feed productivity gains, job growth and broader economic strength.

The scale of the current buildout is already significant. The US had 5,427 data centers at the end of 2025, with thousands more projects announced or at various stages of construction.

Not all of those projects are certain to make it. Many face scrutiny over whether they make financial sense, especially with high yields on the debt used to fund them.

Why a slowdown is on the table

Hassett’s comment about thriving without data center growth lands against a shifting political backdrop. Local resistance to large data center developments has become more pronounced, even as the administration advocates for faster federal permitting.

Changes in state leadership have raised concerns about potential moratoriums on new data center construction.

Geography matters here. Texas alone accounts for about 25% of the nation’s AI infrastructure pipeline, according to Goldman Sachs.

That concentration means policy decisions in a single state could ripple through national investment figures.

What this means

The most interesting part of Hassett’s remark is the hedge built into it. An administration that has leaned on AI investment as a growth catalyst is now signaling that the economy has other legs to stand on.

For investors in companies tied to AI infrastructure, the stakes are concrete. Chipmakers, construction firms, utilities and data center operators have all benefited from the assumption that the buildout keeps accelerating.

There is also a tension between the two halves of Hassett’s message. If AI spending really approaches 3.63% of GDP annually, a meaningful slowdown in that spending would be hard for the wider economy to ignore entirely.

The 4% growth outlook already carries a caveat about external shocks. A sharp pullback in data center construction would not be an external shock, though. It would be a domestic one, shaped by state politics and credit markets.

State-level moves on data center moratoriums, especially in Texas given its share of the pipeline, will signal how much of the planned capacity actually gets built.

Financing conditions matter just as much. If debt yields stay high, more projects could face the kind of viability questions that are already hanging over parts of the sector.