The billionaire chairman of Chinese conglomerate Alibaba Group Holding Ltd. says a bubble is forming in the data center sector.

Joe Tsai said the rush to build the data center infrastructure to power the mass adoption of artificial intelligence is creating a risk of oversupply globally and specifically in the United States. The pace of construction, he argued, is well ahead of any realistic demand projections.

“I start to see the beginning of some kind of bubble,” Tsai said at the HSBC Global Investment Summit in Hong Kong Tuesday, according to Bloomberg. “I start to get worried when people are building data centers on spec. There are a number of people coming up, funds coming out, to raise billions or millions of capital.”

Tech firms, developers and governments around the globe have poured well over $1T into building the infrastructure for the nascent technology of generative AI. Business executives and analysts say it has great potential to increase efficiency and profits, but the tech remains unproven and under development.

President Donald Trump has made AI a cornerstone of his economic agenda, with Apple CEO Tim Cook announcing a $500B investment in data centers after a visit to the White House. That followed a $100B commitment from OpenAI, SoftBank and Oracle that was rolled out in a press conference with the president.

The president’s oldest sons have even gotten into the business, with Donald Trump Jr. and Eric Trump partnering with New York-based investment firm Dominari Holdings to launch a new firm called American Data Centers Inc.

Alibaba, best known for its online marketplace, is investing heavily in AI. The firm plans to spend $52B over the next three years to power its own Qwen large language model and AI platform.

Tsai said the potential for data center overdevelopment was particularly acute in the U.S.

“I’m still astounded by the type of numbers that’s being thrown around in the United States about investing into AI,” Tsai said.

“I don’t think that’s entirely necessary,” he added. “I think in a way, people are investing ahead of the demand that they’re seeing today, but they are projecting much bigger demand.”

Tsai isn’t the first to raise concerns about a potential AI-related bubble. Analysts at TD Cowen raised the flag in February after Microsoft scrapped some of its planned data center leases.

Goldman Sachs Head of Global Equity Research Jim Covello also caused hand-wringing across the sector in June when he published a report questioning whether developers, tech firms and investors were getting ahead of demand.

Firms have spent well over $1T on building out AI infrastructure, but Covello questioned what problem that new capacity is meant to solve.

“AI bulls seem to just trust that use cases will proliferate as the technology evolves,” Covello told Bisnow in July. “But 18 months after the introduction of generative AI to the world, not one truly transformative — let alone cost-effective — application has been found.”