Alibaba Group Chairman Joseph Tsai has raised red flags about a potential bubble forming in AI infrastructure, warning that a global rush to build data centers may be outpacing actual demand for artificial intelligence applications.

Speaking at the HSBC Global Investment Summit in Hong Kong on Tuesday, Tsai said the surge in AI-linked data center construction by tech giants, private equity firms, and other players appears increasingly “indiscriminate.”

“I start to see the beginning of some kind of bubble,” Tsai said, according to remarks reported by Bloomberg.
“I’m still astounded by the type of numbers that’s being thrown around in the United States about investing into AI.”

Bubble Concerns Grow Amid Data Center Race

Tsai’s comments come at a time when major firms like Microsoft (NASDAQ:MSFT), Google (NASDAQ:GOOGL), Meta (NASDAQ:META), Amazon, and even OpenAI have pledged hundreds of billions of dollars to build and expand AI infrastructure across the globe.

But Tsai cautioned that many of these data center projects lack a clear customer base and are being built without a realistic assessment of near-term demand.

“Several projects do not have clear customers in mind,” Tsai noted.

Shift From Training to Inference Reducing Demand

The warning also aligns with broader industry trends showing a shift from AI training to inference, which requires fewer processing resources. As AI models mature and stabilize, the need for expansive, compute-heavy training infrastructure is expected to decline.

This shift was underscored by the success of Chinese startup DeepSeek, which recently launched an AI model matching the performance of top competitors — while using older hardware and lower-cost infrastructure.

Is There Too Much AI Infrastructure?

In February, TD Cowen reported that Microsoft had canceled some U.S. data center leases, adding fuel to speculation about a looming AI infrastructure glut. That has led some analysts to warn that tech companies may have overcommitted resources in the race to dominate the AI space.

Meanwhile, SoftBank and OpenAI last month announced an ambitious $500 billion AI infrastructure initiative in the U.S., though key funding sources remain unclear.

Who’s Benefiting?

So far, the real winners of the AI infrastructure boom have been chipmakers like Nvidia (NASDAQ:NVDA) and TSMC (NYSE:TSM), which supply the high-performance processors fueling AI applications. In contrast, software firms have seen limited near-term gains, despite heavy investment.

Bottom Line

Joseph Tsai’s cautionary words come as the global AI arms race heats up. While AI innovation continues to advance rapidly, his warning suggests that unchecked infrastructure spending, without grounding in real market demand, could signal early signs of an investment bubble.

As AI infrastructure scales rapidly, Tsai’s remarks serve as a timely reminder: Not all growth is sustainable—and not all data centers are needed.

Disclosure: This article does not represent investment advice. The content and materials featured on this page are for educational purposes only.

Related: Trump’s Tariff ‘Liberation Day’ Is 9 Days Away: What You Need to Know as Reciprocal Tariffs Loom

Binance CEO says Trump has been ‘fantastic’ for crypto

North Korea launches new unit with a focus on AI hacking

White House Says Gold Reserves May Be Used To Purchase Bitcoin

Gen Z To Become Richest Generation By 2035: Report

Drama Over Quantum Computing Future heats Up

Nvidia Death Cross Has Traders on High Alert as Momentum Withers

Musk: Tesla to build 5,000 Optimus robots in 2025, tells staff not to sell shares

Trump Promises ‘Flexibility’ on Reciprocal Tariffs

MicroCloud Hologram Reports Strong 2024 Financial Results, Eyes Global Expansion

$1M later, here are the takeaways from sponsoring B2B influencers