Asger Risom
Veteran investor Bill Gurley warns the AI boom is showing classic signs of a bubble, driven by massive spending and rapid wealth gains. He says a market reset is likely as companies struggle to sustain current investment levels.
One of Silicon Valley’s most respected investors says the AI boom may be heading for a sharp correction — and possibly soon.
Benchmark partner Bill Gurley believes the massive surge in wealth and spending tied to artificial intelligence is showing classic signs of a bubble, warning that a reset across the industry is becoming increasingly likely.
A familiar pattern from past tech booms
According to Gurley, the current AI frenzy is following a pattern seen many times before in tech.
As early winners make huge gains, more companies and investors rush in hoping to replicate that success. That wave of enthusiasm often leads to overspending, inflated valuations, and eventually a correction.
“When people get rich quickly, others rush in to do the same,” Gurley said, pointing to that dynamic as a key driver behind market bubbles.
In 2025 alone, the world’s 500 richest people added $2.2 trillion to their wealth — much of it tied to AI. For Gurley, that kind of rapid growth is a warning sign, not just a success story.
The real concern: unsustainable spending
The bigger issue, he argues, is the sheer amount of money being poured into AI.
Tech giants are investing heavily in data centers, infrastructure, and computing power needed to train and run advanced AI models. Analysts estimate that spending could reach around $2 trillion between 2026 and 2028.
That level of investment is already rivaling — and in some cases exceeding — the spending seen during the dot-com boom.
Some companies are also committing to huge long-term costs through data center leases that may not even appear clearly on their balance sheets yet.
For Gurley, the concern is simple: at some point, the money could run out.
Even the biggest players are burning cash
The scale of spending becomes even clearer when looking at individual companies.
Estimates suggest OpenAI could require more than $200 billion in funding by 2030, while total cash burn across the company could reach around $280 billion. Anthropic has reportedly spent over $10 billion training models while generating only about half that amount in revenue.
Gurley compared those figures to Uber’s early years, when a $2 billion annual loss already felt risky.
By comparison, today’s AI spending levels are on an entirely different scale.
Layoffs and “AI efficiency” claims
At the same time, companies are increasingly pointing to AI as a reason for layoffs.
Some executives have suggested the technology could significantly reduce the need for human workers, especially in areas where automation is improving quickly.
Gurley is skeptical of those claims. He argues that companies often use AI as a convenient explanation for job cuts, rather than admitting to overhiring or poor planning.
In many cases, layoffs may be more about managing costs after heavy investment than about AI replacing workers outright.
Sources: CNBC interview with Bill Gurley; Morgan Stanley analysis; HSBC estimates; Moody’s Ratings; Fortune reporting
