TL;DR

A professor sounds the alarm

Scott Galloway, professor at NYU Stern School of Business and bestselling author, has long been one of the most cited voices in technology and markets. Now he is turning his attention to what he describes as a historically dangerous concentration of speculative capital in artificial intelligence (AI).

In an analysis covered by Yahoo Finance, Galloway warns that US equity markets could crash within the next 24 months — and that AI companies could lose between 50 and 70 percent of their current valuations.

"I think the American market crashes because 40 percent of the S&P 500 is now directly or indirectly tied to this enormous bet America is making on AI." — Scott Galloway
Galloway: The AI bubble could crash Wall Street within two years - Bilde 1

Dangerous market concentration

Galloway's core argument centers on vulnerability through concentration. According to him, as much as two-fifths of the S&P 500 index's total value is now exposed to the AI investment wave — either directly through technology companies, or indirectly through supply chains and supporting infrastructure.

The accompanying capital flows are staggering. According to the research material, the world's largest technology companies are expected to spend a combined $700 billion on AI-related capital expenditures in the current year. Historically, capital expenditures at this level — above 2–3 percent of GDP — have preceded market corrections, Galloway notes.

$700bn
Big Tech AI capex 2026
40%
Share of S&P 500 exposed to AI
150x
Market's valuation of AI revenues
Galloway: The AI bubble could crash Wall Street within two years - Bilde 2

Valuations detached from reality

A central part of Galloway's thesis is that the market is pricing AI revenues at 150 times sales — a level he describes as detached from fundamental values. By comparison, Microsoft, Alphabet, and Amazon were priced at roughly 10, 5, and 4 times sales respectively in the period before AI euphoria took hold.

This pricing is driven, in Galloway's view, by what he calls "bubble psychology" — where investors buy in the hope of selling to an even more optimistic investor, the classic "greater fool" mechanism.

Is history repeating itself?

Galloway draws parallels to previous technology bubbles: the railways of the 19th century, electrification, the dot-com bubble, and the telecom crash. What these share in common is that the underlying technology proved genuinely revolutionary — but was quickly commoditized into infrastructure without creating lasting, concentrated shareholder value.

He believes AI could follow the same pattern: a technology that changes the world, but whose gains are distributed across society rather than accumulating in the balance sheets of a handful of companies.

The comparison to the dot-com bubble is explicit. In 2000, the ten largest stocks traded at 52 times forward earnings. Today, what he calls the "Mag 10" — the Magnificent 7 plus AMD, Broadcom, and Palantir — trade at 35 times forward earnings. Lower than the peak in 2000, but still a level that has historically signaled elevated risk.

The ROI problem

One of the most concrete arguments in Galloway's analysis concerns the lack of return on AI investments. According to the research material, he claims that 95 percent of businesses' AI spending cannot be linked to any measurable revenue that a CFO could name.

To justify current valuations through productivity gains alone would require 5–7 million layoffs among the 75 million workers considered vulnerable to AI automation — a scenario Galloway characterizes as the destruction of 10 percent of the workforce.

In addition, internal data from ChatGPT, analyzed by Galloway's team, shows that the share of work-related queries fell from 47 percent in 2022 to 27 percent in 2025. This indicates that AI is in practice being used more for personal purposes than to transform the workplace — which undermines one of the key justifications for the high valuations.

Uber is reported to have exhausted its entire 2026 AI budget before April — and found the technology more expensive than the humans it was meant to replace.

Important caveats

It is worth noting that the research material on which this article is based does not include concrete counterarguments from other economists or market participants. Galloway's analysis is therefore presented on its own terms, without the counterweight that a full debate would provide.

Several analysts have historically warned of AI bubbles without this materializing into market downturns. Galloway is also no neutral observer — he is a prominent commentator with a strong editorial profile, and his analyses should be read in that context.

Nevertheless, the signals he points to — extreme market concentration, high multiples, and a lack of documented ROI — are factors that serious investors should keep on their radar in a market characterized by risk aversion, something today's Fear & Greed Index reading of 31 out of 100 also reflects.

Source: Yahoo Finance / Scott Galloway analysis via Prof G Media