AI Spending Surge Faces Funding Crunch and Recession Risk, Warns Burry and Fitch

AI infrastructure spending is expected to total roughly $10 trillion and average about 3.6% of GDP a year through the early next decade, according to a 247 Wall St. analysis, with the five largest hyperscalers investing nearly $194 billion in a single quarter . Latest-quarter capex ran $54.21 billion at AMZN, $44.92 billion at GOOGL, $35.80 billion at MSFT, $30.12 billion at META and $28.5 billion at ORCL. The piece argues the buildout is so far largely funded from operating cash flow, with hyperscalers reinvesting about four-fifths of it, rather than debt, while flagging strain in Meta's free cash flow of $784 million and negative free cash flow at Amazon and Oracle.

The funding question is where the warnings concentrate. A separate commentary argues the AI boom's biggest risk is financing drying up, not the technology failing, drawing parallels to the dot-com and 2008 cycles when capital for cash-hungry companies disappeared. Fitch modeled a downside scenario in which a 35% fall in U.S. equities and a 15% drop elsewhere would shrink U.S. GDP by 0.6% in 2027 and push global growth below 1%. That is a stress case: Fitch's baseline was raised to 2.6% global growth for 2026.

Michael Burry added a technical critique, arguing that large language models may never reach general intelligence and that training on AI-generated data risks model collapse. For investors in the hyperscalers and chip suppliers, the practical markers are free cash flow trends in the next round of earnings, off-balance-sheet financing for data centers, and whether contracted backlogs such as Microsoft's $678 billion and Oracle's $664 billion keep converting to revenue.