Michael Burry "AI Investment Craze Will Collapse Like 2008"

▲ US stocks, Artificial Intelligence (AI), bear market/AI-generated image

Michael Burry, the investor who predicted the 2008 financial crisis, has once again sounded the alarm on the artificial intelligence (AI) investment craze. The basis for this is the forecast that data center capital expenditure could soar to 3% of the US Gross Domestic Product (GDP). An analysis suggests that if AI demand falls short of expectations, the pace of investment reduction could be comparable to past housing market collapses.

According to the investment specialized media StockTwits on August 7 (local time), Burry focused on three charts released by Torsten Slok, Partner and Chief Economist at Apollo Global Management. Slok analyzed that cloud companies' capital expenditure grew from 0.3% of US GDP in 2019 to 1.4% in 2025. He estimated it would reach approximately 3% of GDP annually from 2027 to 2029.

The pace of AI data center investment is faster than past telecommunications infrastructure investment crazes. In 2000, telecommunications and fiber optic investment peaked at 1.2% of GDP before rapidly shrinking. In contrast, housing investment rose to 6.6% of GDP in 2005, a larger scale than the projected data center investment. Slok stated, "Data center investment, in terms of scale, is smaller than housing, but its increase in proportion to GDP is larger, and its speed is faster than the two past cycles."

Slok warned that investment contraction could have a greater impact on the economy than investment expansion. US housing investment fell from 6.2% of GDP in early 2006 to 3% by the end of 2008. He said, "A cycle that expanded by 0.85 percentage points annually can reverse at a similar speed," adding, "When AI demand falls short of expectations, the real macroeconomic risk is not the investment expansion itself, but the subsequent contraction process."

Big Tech's investment scale is already growing to astronomical levels. The combined capital expenditure of Amazon (AMZN), Alphabet (GOOGL), Meta Platforms (META), Microsoft (MSFT), and Oracle (ORCL) is estimated to be approximately $800 billion this year. A significant portion of this investment is expected to be allocated to the construction of new data centers. S&P projected that AI investment by US hyperscalers could reach $1 trillion by 2027. Bank of America estimated the investment scale for 2026 to be $800-$900 billion and predicted it would exceed $1 trillion in 2027.

Peter Diamandis, Executive Chairman of the X Prize Foundation, claimed that 2,441 data center projects are scheduled to commence in the US from 2026 to 2028, with planned investments totaling up to $2.48 trillion. However, he did not disclose the source of these figures. Burry continues to maintain that AI infrastructure investment could eventually lead to oversupply. Since late 2025, he has established or expanded bearish positions on Nvidia (NVDA), Palantir (PLTR), Caterpillar (CAT), Applied Materials (AMAT), Micron (MU), Oracle, and Nebius (NBIS), among others.

[Article Key Summary]

-Michael Burry highlighted an analysis suggesting that AI data center capital expenditure could grow to approximately 3% of US GDP between 2027 and 2029.

-Torsten Slok warned that if AI demand falls short of expectations, investment expanded at an annual rate of 0.85 percentage points could contract at a similar speed.

-Major US Big Tech companies' capital expenditure this year is estimated at around $800 billion, with projections that it could exceed $1 trillion by 2027.

*Disclaimer: This article is for investment reference only, and we are not responsible for any investment losses based on it. The content should be interpreted for informational purposes only.*