Justin Hotard, the chief executive of Nokia, said in an interview published on Monday that tech companies would build data centers twice as fast if shortages of memory chips and energy weren’t holding them back.

Mr. Hotard, speaking on CNBC’s The Tech Download podcast, rejected the idea that the industry was building too much. “I don’t think you can say in any manner we’re overbuilding today because reality is that if we could build 2x faster, our customers could build 2x faster, they probably would,” he said.

“So that gives me confidence that we’re still in the early days,” he said of the AI buildout.

His comments came weeks after a scare over AI safety. An Anthropic researcher’s warnings that the technology posed an existential threat led Dario Amodei, the company’s chief executive, to call for a slowdown in work on the most advanced models, and AI stocks sold off last month, according to CNBC.

Mr. Hotard said demand didn’t depend on labs like Anthropic and OpenAI putting out new models. “I think the reality is that we’re even where we are today is so early in the deployment, and the power of these models is so massive that there’s a lot of there’s a lot of demand to run in just deploying,” he said.

“Even if we didn’t have another frontier model released in the next three years, we could probably make tremendous progress just deploying the technology that’s there today.”

Nokia, the Finnish telecommunications equipment maker based in Espoo, doesn’t sell the chips. It makes the gear that connects racks of them inside a data center and links data centers in different places. Sales to AI and cloud customers rose 105 percent to about 446 million euros in the second quarter, [according to the company’s results](<https://www.nokia.com/newsroom/nokia-corporation-report-for-q2-and-half-year-2026/>), out of total sales of 4.82 billion euros. Those customers placed 2.8 billion euros in new orders in the quarter, [Reuters reported](<https://live.euronext.com/en/financial-news/nokia-q2-profit-beat-sales-ai-cloud-doubled>).

Mr. Hotard said in July that supply, not demand, was the industry’s main constraint, and that it was pushing customers to sign longer-term orders. He ran Intel’s data center and AI group before Nokia hired him in 2025.

Investors have bought in. Nokia’s shares are up about 130 percent over the past year.

The shortages he named have reached consumers. DDR5 memory prices in Germany are up 414 percent in a year, and memory can make up as much as half the cost of parts in a budget phone, according to The Next Web.

Not everyone is sure the money will be there. In [a paper for the Brookings Papers on Economic Activity](<https://www.brookings.edu/articles/financing-the-ai-buildout/>), presented on Sept. 25, Stijn Van Nieuwerburgh, a professor of real estate at Columbia Business School, estimated that investment in data centers, chips, power and networking would total $10.3 trillion from 2025 to 2032. That is an average of 3.63 percent of U.S. gross domestic product a year, bigger relative to the economy than the canal, railway, electrification, highway and telecommunications booms, he wrote.

Dr. Van Nieuwerburgh also wrote that the financing was moving off companies’ balance sheets and into joint ventures, private credit and special purpose vehicles. Those deals only work if demand for AI services holds for years.

Others in the business sound like Mr. Hotard. CoreWeave’s chief executive said last week that a backlash against data centers hadn’t touched orders.

Bain put a number on the gap last week, according to The Next Web. It said AI would need $6 trillion a year in revenue by 2031 to pay for the data centers now under construction. Today’s products could bring in $1.2 trillion to $1.8 trillion.