Driving the news: A significant gap is growing between AI infrastructure spending and expected returns.
- While AI vendors are expected to spend $7 trillion on infrastructure by 2030, Gartner research projects that they will generate no more than $3.6 trillion in related revenue over that same period.
- More than $4 trillion in AI infrastructure spending could be at risk of depreciation by the end of 2032 if companies maintain their current pace of investment.
- Gartner also predicts that more than 80% of specialized AI cloud providers, called neoclouds, will cease to operate independently by 2030.
Concrete example: CoreWeave, one of the largest neoclouds with some of the closest ties to Nvidia – the financial and spiritual foundation of the AI boom – has been facing a sell-off in its stock and debt for several days.
- Returns topped 15% and its shares fell 15.5% before rebounding after Nvidia CEO Jensen Huang spoke on the company’s behalf.
Follow the money: Wall Street sources told Axios that as long as AI development was paid for in cash, there was nothing to worry about. But today, a third of investments in AI come from debt, and financial markets are showing signs of severe fatigue.
- It is increasingly difficult to obtain financing. Global AI-related debt issuance plunged to $23 billion in September 2026, down sharply from $113 billion in June, according to the Financial Times.
- Banks are actively withdrawing to avoid concentration risk, with JPMorgan CEO Jamie Dimon confirming his bank is tightening credit conditions.
- Infrastructure projects and providers are in trouble: About $18 billion in bank loans for an Oracle data center campus recently slipped into a stress zone, while Nvidia-backed AI cloud provider Firmus suspended an IPO that sought a valuation north of $30 billion.
Zoom: Constant spending is based on the belief that demand for computing will never wane – a belief that has structural cracks.
- Ioannis Antonoglou, co-founder of Reflection AI, told Semafor that companies may need “a decade or two decades” to fundamentally rethink their operations around AI.
- It’s been 10-20 years of companies slowly adopting the technology that is being built (expensively) today.
Between the lines: This creates a disconnect: Data centers and chips require huge investments today, but the returns that could justify those expenses will take years to materialize.
- Simply put, there is a huge gap between AI spending and revenue, one that will not close for many, many years.
- At the same time, companies that are rapidly adopting AI are using the tools more and more efficiently, turning to less compute-intensive open models to reduce costs.
Reality check: Look to Anthropic’s latest financial data for a counterpoint to the declines.
- The company is accumulating cash, making its free cash flow positive in the third quarter and increasing its revenue by almost 50% compared to the April-June period.
- But even Anthropic’s staggering growth has slowed, although that’s to be expected because trees can’t grow to the sky.
The bottom line: Debt-fueled AI development is not an isolated issue that only tech investors need to worry about.
- The stock market and economy are increasingly linked to the AI boom.
- If the AI gamble doesn’t work, markets and the economy could also falter.
Gn bussni
