This story was originally published on MyNorthwest.com.
The global investment in artificial-intelligence companies has soared to roughly $500 billion, fueling a surge in valuations and sparking comparisons to past tech manias like the dot-com era.
However, some analysts argue this may all be “one big AI bubble,” pointing to three different reasons — from circular investment to overvaluation of AI companies’ overall worth. Let’s break it down.
Circular financing
Tech companies investing in other tech companies is fueling much of the boom. Firms, from Microsoft to Meta, and Amazon to OpenAI, Anthropic, and others, are increasingly handing over billions of dollars to prop each other up in ways that many analysts say are resulting in synthetic growth, while organic growth and real customer demand for actual products have slowed.
For instance, one major player recently poured $5 billion into a leading AI startup, which, in turn, committed tens of billions for cloud services from another tech giant, while a third prominent server and chip maker invested billions more into an AI startup. Critics worry cycles like this inflate valuations without delivering real growth.
Credit vs cash
While many of those cyclical investment deals are paid in cash, many more are paid on credit, and credit-fueled deals raise the stakes. Especially for smaller AI companies, if one of the spokes in their investment wheel defaults, there’s a higher risk of financial disaster.
Looming accounting questions
Accounting practices and depreciation concerns cloud the picture. Some financial veterans allege companies are underreporting future costs. For example, stretching out depreciation on costly AI chips over many years, while the hardware becomes obsolete much faster.
“If you say a chip lasts six years instead of three, you only expense half as much each year, which makes your profits look bigger,” explained an analyst.
One prominent investor has even placed big bets against leading AI firms, citing these concerns. Michael Burry is an American investor and hedge fund manager who might be best known for being among the first investors to predict and profit from the subprime mortgage crisis in 2007. Claiming there are parallels between the housing market and the AI tech market, including overvaluation, he is now betting against many AI companies.
Adding fuel to the Burry factor, a new report from the Massachusetts Institute of Technology delivers a stark conclusion that 95% of AI pilot projects inside large companies fail before they ever get off the ground. While executives often point to regulatory hurdles or still-maturing technology, MIT researchers said the deeper problem is a lack of integration and imagination.
According to the report, off-the-shelf AI tools like ChatGPT may impress individual users with their flexibility, but inside corporate workflows, they often fall flat because, according to the report, they are poorly aligned, inadequately trained, and ultimately abandoned.
Adding even more fuel to those factors, Deutsche Bank is warning that the current AI boom can’t last forever. The bank said tech spending is growing so fast that it’s actually keeping the U.S. out of a recession, but warned that the pace won’t continue.
“I think the AI bubble is a bubble, and it will pop,” one Deutsche Bank analyst explained.
At the same time, proponents of the AI revolution point out that the growth isn’t all hype. According to recent data, there are now nearly 500 “AI unicorns” or private AI companies valued at $1 billion or more. Their collective worth is roughly $2.7 trillion. Many of them have emerged only in the past two years. That rapid scale-up reflects strong investor confidence and, supporters said, long-term potential.
And some market leaders insist the infrastructure being built now, from chips to data centers, will underpin years of real AI adoption, not just speculation.
As the dust settles, the big question remains — is this the start of a transformative era, or the calm before a dramatic crash? And if it does turn into a dramatic crash, another question looming is who and what will be responsible for picking up the pieces, especially given the race between the U.S. and China to lead AI markets?
Sam Altman, CEO of OpenAI, which developed ChatGPT, recently said in an interview that, given the potential national security stakes, just like during the housing market failure, some AI companies may be too big to fail.
“Given the magnitude of what I expect AI economic impact to look like, sort of I do think the government ends up as like the insurer of last resort,” Altman said.
Follow Luke Duecy on X. Read more of his stories here. Submit news tips here.
©2025 Cox Media Group
