SerpApi.Org
The artificial intelligence spending spree sparked by the release of ChatGPT 3.5 in November 2022 has been unprecedented in its scale.
Global AI infrastructure spending is expected to reach $758 billion by 2029, according to the International Data Corporation. That’s more than double the $300 billion spent in 2025.
“The AI bubble isn’t a result of any actual return on investment,” frequent AI critic Ed Zitron wrote in a recent scathing takedown of the industry. He called the outsized AI spending and the economic bubble it has inflated "the greatest capital misallocation in history."
ChatGPT parent company OpenAI has already declared that it intends to spend more than $852 billion by the end of 2030. It intends to spend $50 billion or more on compute power just this year, which, according to Zitron’s math, is more than 50% of all global AI compute spend.
For comparison, Chipotle's entire market cap is about $42 billion. But Chipotle also generated about $11 billion in revenue last year, compared to OpenAI's 13.1 billion in revenue.
That incongruity was behind some of the tech sell-off Wall Street witnessed last week.
Big tech AI spending runs into investor resistance
Last week, investors made major pullbacks in chip stocks, leading to a broader sell-off in tech stocks. Information technology was the worst-performing group in the S&P 500 Index last week, falling 1.6%. Meanwhile, the tech-heavy Nasdaq 100 Index also lost 4.1%.
Investor concern about runaway AI spending with little return on investment is behind the selloff, analysts told Bloomberg this week.
“Investors are getting to the point where they’re uncomfortable with how much money is being spent, and they’re worried about a bubble,” said Jake Seltz, portfolio manager at Allspring Global Investments. “Ultimately, we need to see a re-acceleration in revenue.”
OpenAI has the most popular consumer-facing AI product, with ChatGPT boasting more than 900 million weekly users. But even as the company's revenue has jumped from $2 billion in 2023 to a projected $29 billion in 2026, the company is also losing market share and money on every new user that utilizes its platform.
OpenAI lost about $5 billion in 2024; it expects to lose about $14 billion this year, but outside watchers estimate it will lose as much as $36.6 billion in 2026.
“OpenAI’s collapse will be a direct result of its loss-laden economics — its doomed, loss-making subscriptions, its pathetic advertising revenue, and API costs that became a ‘huge issue’ for its enterprise customers — and the fact that outside of the hype, AI lacks measurable ROI,” Zitron says.
But OpenAI is still a private company, so it doesn't have to worry about retail investor sentiment just yet. The same can't be said for some of its biggest backers.
Alphabet is scheduled to report second-quarter earnings on Wednesday, July 22, and investors are selling heading into the print. The stock rallied on Monday, July 20, climbing more than 2.5% at last check, but the stock dropped more than 6% over the previous two sessions last week.
Microsoft is coming off its worst month since 2000, according to Bloomberg, and the stock has lost nearly 20% year to date.
“At some point, earnings are being questioned so much that you can’t put as high of a multiple” on these stocks, Todd Ahlsten, chief investment officer at Parnassus Investments, told Bloomberg. “There’s going to be a lot more focus on cloud gross margins, pricing, what kind of AI revenue is being generated per dollar of compute.”
Amazon, Alphabet, Meta and Oracle are known as hyperscalers due to the hundreds of billions they've committed to building out AI and AI infrastructure. But that spending has decimated their free cash flow, with 2026 estimates for the group dropping to nothing from $300 million in 2024.
Investors urged to be more careful about AI stocks
Investors piling into AI-related stocks helped drive the stock market to all-time highs in June, and while there are still companies with attractive balance sheets, there are fewer than there used to be.
“At some point, earnings are being questioned so much that you can’t put as a high of a multiple” on these stocks, said Todd Ahlsten, chief investment officer at Parnassus Investments. “There’s going to be a lot more focus on cloud gross margins, pricing, what kind of AI revenue is being generated per dollar of compute.”
One of the most glaring signs that investors are beginning to tire of the unfettered spending on AI is Apple. Apple has avoided large capital expenditures on AI, instead opting to partner with model providers to power its AI services, and has been the best Magnificent 7 performer by a wide margin with a 23% gain so far in 2026.
“The existence of OpenAI justified an era of mania and opulence. Hyperscalers, bereft of new hypergrowth ideas, were able to point at the fact that ChatGPT had ‘the fastest growing user base of all time‘ and the Microsoft ‘supercomputer’ that built it and tell their investors that if they didn’t invest, they’d be left behind, with Amazon, Meta, and Google announcing their own nebulous ‘supercomputers’ in 2023,” Zitron said.
“This is the underlying greed that has driven this wasteful, reckless and destructive era — the belief that there will be another OpenAI and, as I’ve said, the chance to become the next OpenAI’s landlord,” Zitron said. “And like any great investment bubble, the more money that piled in, the greater the fear of missing out, the more dollars that can be justified in turn, and the more complex and deranged the mythology becomes.”
