The Race for AI Data Centers Could Lead to a Bubble

President of Alibaba Group Holding, Joe Tsai, warned of the possibility of a bubble due to the construction of data centers, emphasizing that the pace of this development could outstrip initial demand for artificial intelligence services. Speaking at the HSBC Global Investment Summit in Hong Kong, Tsai highlighted that a race is underway among technology companies, investment funds, and other entities to build data centers – from the U.S. to Asia – but many of these projects lack clear customers.

According to data from Synergy Research Group, global spending on data centers reached a record $250 billion in 2024, with an expected further growth of 15 percent annually. However, analysts are increasingly warning that the rapid increase in capacity could result in underutilized resources, especially in regions where current demand does not justify high investments.

Tech Giants Invest Billions

Large technology companies, including Microsoft and SoftBank, are investing billions of dollars in purchasing chips from Nvidia and SK Hynix, which are crucial for AI development. According to Gartner’s research, global demand for AI chips has increased by 50 percent in the last two years, with Nvidia holding over 80 percent of the market for AI GPUs.

Alibaba, which announced in February that it is fully directing its strategy towards AI, plans to invest over 380 billion yuan ($52 billion) in the next three years. At the same time, in the U.S., former President Donald Trump is promoting the Stargate project, which envisions an investment of up to $500 billion in AI infrastructure, including data centers, optical networks, and computing capacities needed for the development of advanced AI models.

Investor Concerns Are Growing

Following Tsai’s statements, Alibaba’s shares fell by more than 3 percent on the Hong Kong stock exchange, while shares of other Chinese tech companies, such as Tencent Holdings and Baidu, also recorded declines.

On Wall Street, concerns are rising over excessive spending on AI infrastructure, especially after the Chinese company DeepSeek introduced an open-source AI model that, according to their claims, competes with American technology at significantly lower costs. Critics also warn of a lack of concrete, real applications of AI beyond content generation and data analytics.

– I am beginning to notice signs of a bubble. We see projects being launched without secured users. It is concerning when data centers are built speculatively, and investment funds and other entities are raising billions of capital without a clear strategy – Tsai told summit participants.

Alibaba aims to regain market dominance by 2025, with the AI platform Qwen playing a key role in this process, which is expected to strengthen its core e-commerce segment and cloud services. Tsai noted that the company is in a ‘reboot’ phase after years of regulatory oversight that limited growth, and is currently intensively hiring AI experts to realize its vision of developing artificial general intelligence (AGI).

American Tech Companies Under Scrutiny

Tsai criticized the massive investments of American tech giants in AI infrastructure. Amazon, Alphabet, and Meta Platforms have announced investments in AI this year amounting to $100 billion, $75 billion, and up to $65 billion, respectively. However, analysts at TD Cowen reported in February that Microsoft canceled some data center lease agreements in the U.S., raising doubts about whether the company will need such a large amount of AI capacity in the long term.

Microsoft downplayed the concerns, emphasizing that they are spending more than ever in their history, primarily on chips and data centers. The company estimates that it will invest $80 billion in AI infrastructure this fiscal year, but the pace of growth of these investments is expected to slow down from July 2025.

– I am astonished by the figures being discussed in the U.S. regarding investment in artificial intelligence. They literally mention hundreds of billions of dollars, even $500 billion. I think that is not entirely necessary. In a way, investments are outpacing current demand, assuming much greater demand in the future – concluded Tsai.