If there’s an AI bubble, this is what could go wrong
Marketplace®
Daily business news and economic stories
Donate
Volume
00:00
-- : --
listen
[undefined]
--------- ---- ---- ---- ---- ---- -------- ---- ---- ----
Feb 23, 2026
What would the stock market and economy look like without AI?
AI has helped boost the stock market and GDP. But the prospect of a bubble has some experts concerned about the technology’s influence.
Technology
by
Janet Nguyen
The Magnificent 7 has helped boost financial markets, but too much exposure to top-performing tech companies means your 401(k) could become vulnerable.
tadamichi/Getty Images
Last year, the Magnificent 7 grew 23%. The rest of the S&P 500? Only about half of that, growing at a rate of 11%.
The Magnificent 7 — a group of tech companies with an outsize influence on the stock market — include Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla.
These companies have surpassed the gains of other companies in the S&P 500, or the
“S&P 493,”
in recent years, according to data that FactSet, an analytics firm, provided to Marketplace.
These top tech companies, which have invested heavily in AI, have helped boost financial markets. Meanwhile, AI-related spending may have helped lift economic growth over the past several quarters, although economists disagree on how much of an impact AI has had on gross domestic product.
In the first quarter of 2025, gross domestic product shrank 0.65%. If you take out AI-related contributions, then GDP ended up shrinking 1.94%, according to data that Hannah Rubinton, an economist at the St. Louis Fed, provided Marketplace.
But some experts have raised the prospect of an AI bubble amid the fast-paced growth of these tech companies and compare it to the dot-com bust in the early 2000s, when online startups crashed after their share prices reached unrealistic levels.
“There are clearly a number of companies that have emerged in this AI moment who, once we take a finer look at, probably have suspect business models and suspect prospects that might be overvalued now,” said Michael Lenox, a professor of business administration at the University of Virginia. “So I do think we're in a bubble in that sense.”
However, when new technologies emerge, we eventually have what’s called “the shake out,” where there are both winners and losers, Lenox said. He pointed out that while there were companies that failed during the dot-com bust in the early 2000s, there were internet-based companies that became successful. “This is a once-in-a-generation technology that's impacting a wide variety of sectors,” he said.
Lenox said he thinks the winners in the AI race will be those with their own core AI technology, while he questions the future of those who are building products off of other companies’ AI models.
Market returns
Over the past three years, the Magnificent 7’s market performance has outpaced the S&P 493. In 2023, the Magnificent 7 grew more than 33.3%, while the S&P 493 declined 4%. These 7 stocks grew at a slower rate last year, but still had a 10-percentage-point lead over the rest of the S&P.
One bright spot for American workers with retirement accounts: As the markets go up, so do their investments.
There’s always the risk that some companies are overvalued and end up crashing like during the dot-com bust in the early 2000s, said Eric Swanson, an economics professor at the University of California, Irvine.
But there was “a lot more euphoria and sort of a frenzy around dot-com companies,” many of which had no profits at all despite their stock valuations, Swanson pointed out.
Companies in the Magnificent 7 are also not taking out an excessive amount of debt to finance their AI investments and are instead relying on their internal cash flows, said John Bai, a finance professor at Northeastern University.
And companies like Alphabet, Meta and Nvidia have other businesses they can fall back on, said Derek Horstmeyer, a finance professor at George Mason University.
However, Horstmeyer said he’s concerned about circular financing among the Magnificent Seven.
“A circular financing deal would be Nvidia agrees to invest in Oracle, and Oracle agrees to buy Nvidia chips or use Nvidia products. So it’s a two-way street in the deal. Both are putting investments into each other,” Horstmeyer said.
These deals end up bidding up the stock price of other companies within the Magnificent Seven, Horstmeyer said.
“Circular financing can artificially raise the prices of company stocks, but it can go wrong if one of the companies in the chain has problems and cannot meet its obligations,” Horstmeyer said.
So if one company runs into financing issues and can't fulfill its obligations to another company, like buying their products, then the stock price of the second company will fall, causing the share prices of the next linked company to fall as well, Horstmeyer explained.
“For folks with 401(k)s, they undoubtedly have massive exposure to the Mag 7 if they own the S&P 500. So if the Mag 7 falls, then people can lose a lot in their retirement savings,” he said.
GDP growth
AI-related investments in data centers, research and development, software and processing equipment helped boost GDP in 2025, according to Rubinton at the St. Louis Fed.
“The contribution of artificial intelligence to GDP growth in the first three quarters of 2025 was comparable to the height of the dot-com bubble in 2000. AI accounted for 39% of GDP growth in 2025 (through the third quarter) versus 28% in 2000,” Rubinton said over email.
But in the third quarter, AI-related contributions "normalized" relative to the second quarter, Rubinton said. “While AI investments are still high, their quarterly growth rates have tapered off,” Rubinton noted.
Economists have debated how much AI has contributed to economic growth, with some experts
saying that AI has actually had zero impact
and that spending on foreign-made components is helping other countries’ economies, not America’s.
AI could end up boosting labor productivity, which
could in turn lead to higher wages
without leading to higher inflation, but
the link between the two is murky,
Marketplace previously reported.
It’s hard to determine how much AI is boosting productivity vs. factors like improved business practices and a loose job market, said Erika McEntarfer, research scholar with the Stanford Institute for Economic Policy Research, in a previous Marketplace interview.
If companies aren’t hiring, but output grows, then labor productivity will increase naturally, McEntarfer said.
Because companies are in the experimental stages of AI, it might take a while for AI to actually boost productivity, McEntarfer explained.
Not all of these AI investments may pan out, said Steven Kyle, an associate professor of applied economics at Cornell University.
So far, we’ve only seen the economic effects from the investment in AI capacity, Kyle said.
“We haven't had much return at all yet from the actual products or services that AI is supposed to provide,” Kyle said.
Although AI has helped boost GDP growth in recent quarters, if people are laid off because of AI, that could end up hurting GDP growth in the long run, Horstmeyer said.
“The U.S. economy is built on consumption,” Horstmeyer said. “If everyone gets laid off by AI, everyone loses jobs, then consumption is going to go down, and that's going to really hurt GDP.”
Related Topics
Collections:
Technology
Tagged as:
artificial intelligence
stock market
GDP
Our best stories, straight to your inbox
Sign up for the Marketplace newsletter to get stories that help you live smarter — plus analysis, explainers and our song of the week — sent to you every Friday morning.
Latest Stories
Marijuana's reclassification raises new questions for the cannabis industry
What is the informal economy?
Why Florida is no longer the king of citrus
Stories You Might Like
View All
Prediction markets to start offering risky "perpetual futures"
Companies
She landed the job, but she still feels the pressure of an uncertain economy
Workplace/Jobs
When do tech companies need to be consistently profitable?
Marketplace Tech
Latest Episodes
View All Shows
2 hours ago
6:58
4 hours ago
14:20
15 hours ago
25:28
19 hours ago
13:54
a day ago
34:11
3 days ago
32:17