AI Datacenter Bubble? Alibaba's Tsai Raises Red Flags Amid Soaring Investments

Alibaba (NYSE:BABA) Chairman Joseph Tsai has issued a warning about a potential bubble forming in AI datacenter construction, citing the growing disconnect between infrastructure development and actual demand. Speaking at the HSBC Global Investment Summit in Hong Kong, Tsai highlighted a global surge in AI server building by tech giants, private equity firms, and others — often without a clear customer base.

“I start to see the beginning of some kind of bubble,” Tsai said, according to Bloomberg. He pointed out that the pace and scale of AI-related infrastructure projects appear indiscriminate and may outstrip the current need for AI services.

The concerns arise as the AI industry undergoes a shift from training large models to inference — generating results from pre-trained models — which demands less processing power. Meanwhile, Chinese AI startup DeepSeek recently unveiled a competitive model using older, less expensive chips, casting doubt on the necessity of high-end hardware.

U.S. tech giants like Microsoft (NASDAQ:MSFT), Google (NASDAQ:GOOGL), Meta (NASDAQ:META), and Amazon have collectively committed hundreds of billions of dollars to AI infrastructure. However, Microsoft’s reported cancellation of some U.S. datacenter leases, as noted by TD Cowen in February, has sparked fears of overcapacity.

SoftBank (OTC:SFTBY) and OpenAI also announced a $500 billion AI infrastructure plan in the U.S., though funding details remain unclear. Tsai expressed amazement at the scale of investment being discussed in the U.S. market.

While the AI boom has significantly boosted chipmakers like Nvidia (NASDAQ:NVDA) and TSMC (NYSE:TSM), software firms have yet to see proportionate returns, raising questions about the sustainability of current investment trends.