‘AI Bubble’ Alarm Spreads on Wall Street Amid Frenzied Investment Without Clear Revenue Models

Wall Street Warns of Market Correction in AI Investment Boom
JP Morgan CEO: “AI Stocks Already in Bubble Territory”
Massive Capital Concentrated in OpenAI, Anthropic, and xAI

A growing sense of unease is rippling through global financial markets as the artificial intelligence (AI) investment frenzy shows signs of overheating. JPMorgan CEO Jamie Dimon has warned that AI equities have already entered bubble territory, while the International Monetary Fund (IMF) has voiced concern over excessive concentration in AI mega-caps and the potential for a sharp correction in tech stocks. The excessive flow of capital into OpenAI and a handful of startups, coupled with weak profitability, has reignited memories of the early-2000s dot-com bubble and heightened caution among investors about an impending market adjustment.

IMF Warns of Excessive Market-Cap Concentration in AI Giants

According to Yahoo Finance on October 21 (local time), anxiety is building on Wall Street as AI optimism reaches fever pitch. “We are seeing asset prices rise to worrisome levels, with AI equities clearly in bubble territory,” Dimon told reporters. “When prices rise this far, a correction becomes inevitable — meaning the risk of loss is growing.” A global fund manager survey by Bank of America (BoA) found respondents naming an AI stock bubble as the single greatest risk to world markets. Among the roughly 200 fund managers surveyed, the average cash allocation dropped to 3.8 percent — a level typically seen when risk appetite peaks.

While U.S. Treasury Secretary Scott Bessent countered that “AI remains in the early innings,” the IMF has now lent weight to bubble concerns. In its Global Financial Stability Report released on October 14, the Fund warned that “market-cap concentration in AI mega-caps is excessive” and cautioned that “if tech-sector earnings fail to justify lofty valuations, a sharp and disorderly correction could occur.” The report further noted that “with high debt levels and persistent uncertainty around major economies’ monetary policies, a sudden downturn in tech valuations could exert downward pressure on global growth,” adding that “overblown expectations surrounding AI could destabilize markets in the short term.”

AI Revenues Insufficient to Support Computing Capacity Expansion

Warnings of an AI bubble have circulated since early this year, but the tone has shifted in recent months. Whereas mid-year discussions centered on AI infrastructure spending by Microsoft, Google, and Meta, the focus has now moved to the billions flowing into select startups such as OpenAI, xAI, and Anthropic — whose valuations run into the hundreds of billions of dollars. Analysts caution that Big Tech’s infrastructure build-out largely serves the computing needs of these few firms, raising concerns that the broader equity market is being driven by a narrow cluster of capital-intensive players.

Some market strategists are drawing parallels to the early-2000s dot-com bubble. One indicator fueling that comparison: S&P 500 companies are currently trading at about 23 times forward earnings — far above the ten-year average of 18.7 times — levels widely viewed as symptomatic of speculative excess. Anthony Saglimbene, chief market strategist at Ameriprise Financial, observed that “investors are becoming increasingly skeptical about the massive capital-expenditure announcements,” adding that “there is growing anxiety about how much money is being spent and how little of it is being recouped.”

Global consultancy Bain & Company estimated in a recent report that AI companies would need annual revenues of roughly $2 trillion by 2030 to sustain computing-power growth, yet actual revenues are expected to fall short at under $800 billion. Prominent hedge-fund manager David Einhorn of Greenlight Capital echoed the alarm, saying, “The numbers being floated around AI investment are so extreme that they are almost incomprehensible. While a total collapse isn’t guaranteed, there is a very high probability of massive capital losses in this cycle.”

OpenAI’s Returns Fall Below 10 Percent of Investment

At the epicenter of the debate is OpenAI. On October 19, Citigroup analyst Chris Donnelly told clients that OpenAI would need to build computing capacity totaling 26 gigawatts (GW) through its partnerships with global semiconductor firms — requiring at least $1.3 trillion in capital expenditures by 2030. Yet CEO Sam Altman has reportedly outlined an even more ambitious target: according to The Information, Altman aims to secure 250 GW of computing power by 2033, implying capex as high as $12.5 trillion.

The challenge, however, lies in the imbalance between massive investment and meager returns. Citigroup projects OpenAI’s revenue will reach just $163 billion by 2030 — less than one-tenth of its expected capital outlay. Critics argue that AI firms still lack a viable, scalable business model to justify these valuations. The debate was further inflamed by an MIT report titled “The Generative AI Gap: The State of Enterprise AI in 2025.” The 26-page study found that while corporations have invested between $30 billion and $40 billion in generative AI, 95 percent have yet to see any return. Of the nine industries analyzed, only technology and media showed meaningful productivity gains.

The report identified poor learning adaptation as the chief reason AI has failed to boost corporate productivity. A chief operating officer of a mid-sized manufacturer remarked that “the only change we’ve seen from AI adoption is processing some contracts a little faster.” Many enterprise systems, it noted, lack the ability to learn, adapt, and evolve to the unique contexts of individual workplaces. In practice, most AI tools rely on natural-language data, forcing manufacturers dealing with design schematics to repeatedly reprocess and retrain information. “Users still turn to ChatGPT for simple tasks,” the report concluded, “but its limited memory prevents effective use for mission-critical operations.”