Is the AI Bubble About to Burst?

Jon Gingerich

Maybe you’ve noticed something in the last month: Everyone seems to be talking about a bubble in the AI sector. I’ve been hearing whispers about it since the beginning of the year, but for the past few months, the tenor has changed, the chatter palpably louder.

Unless you’ve been living under a rock, you’d know the markets have reached all-time highs this year, powered primarily by ridiculous investments in the tech sector. And a disproportionately large amount of this investment accounts for spending on AI companies as well as the ancillary products and services that provide AI’s infrastructure or its physical materials: everything from data centers to chip makers to the mining companies they rely on to build them.

AI chipmaker Nvidia broke records this year when it became the first publicly traded company to be valued at $5 trillion, which, by comparison, is greater than the GDPs of Germany, India or Japan. The company, which has seen its stock surge 50 percent in the last year, now makes up almost 10 percent of the S&P 500. Software company Palantir’s stock price has gained more than 150 percent in the last year. Nuclear technology company Oklo, which has received massive investor attention due to speculation that its nuclear energy might someday power AI data centers, has seen its stock price shoot up more than 400 percent in 2025. Other big AI players, such as chipmaker AMD, Google, Microsoft and Meta, have seen their stock prices surge in recent months.

You don’t need to be Jim Cramer to realize there’s a lot of eggs being put in one basket here, that too much of the U.S. economy is riding on the success of one sector, that the financial markets are increasingly being held aloft by a few mega-cap AI darlings. I’m no expert, but this sort of concentration can’t be good. Especially considering the tech sector’s big AI bets aren’t resulting in the returns one might expect. Investors are blindly throwing wads of cash at any company containing the letters AI, which, of course, is causing more companies to rush more unbaked AI products into the marketplace, and much of what they’re producing, unfortunately, is underwhelming. Indeed, a damning August MIT report found that 95 percent of corporations that adopted Generative AI programs saw those initiatives fail to deliver any measurable financial value. The hype simply isn’t matching reality; the stock price is superseding whatever underlying value those products actually have. And this kind of speculation over value is an archetype example of what makes an economic bubble.

So, perhaps it’s no surprise that a growing chorus of investors is now warning that current AI valuations are untenable and that the stock market is primed for an imminent correction. Investor and hedge fund manager Michael Burry—who famously predicted the 2008 U.S. housing market crash—made headlines when he implied in a cryptic social media post that the AI bubble was here. Burry’s ominous warning was followed by comments from CEOs of Goldman Sachs and Morgan Stanley, who both suggested we could see a drawdown of somewhere between 10–20 percent in the markets within the year. It was later discovered that Burry has been betting against Palantir and Nvidia, just like how he shorted the subprime mortgage market. That’s not a good sign.

Reality finally seems to be catching up with the markets. Tech stocks took a dive in late October, setting the stage for an unusually sour period. Palantir’s stock plunged, with Nvidia following suit. AMD cratered and Meta’s stock slipped after reporting disappointing third-quarter results, and the same happened to Google and Microsoft despite resilient quarterly earnings. Add to that a weakening labor market, ongoing tariff drama and massive recent layoffs that’ve rocked the tech sector—Amazon, IBM, Intel, Oracle, Salesforce, among others—and we might be compelled to ask: Is the runaway AI bull finally losing steam?

Granted, people have made a lot of predictions about the stock market over the years. There are plenty of reasons to believe the rally could continue and the AI wave will keep rolling, at least for the time being. Especially considering the federal government’s recent 10 percent stake in Intel, as well as President Trump’s proclivity to treat Nvidia like a bargaining chip when discussing trade with China. And the fact remains that, given the enormous levels of current spending on AI infrastructure, investors remain optimistic that AI valuations will remain high. Finally, there’s the unshakable idea that AI is, by definition, one of those rare, once-in-a-generation, “this changes everything” breeds of revolutionary technology. There remains a lot of interest in AI, and that doesn’t appear to be going away anytime soon.

There have been a number of bubbles over the years. We’ve been down this road enough times to know that innovation arises out of ecosystems characterized by incredible waste. Think of the early web players that didn’t survive the dot-com crash: AltaVista, Excite, Lycos, Pets.com, Webvan and countless others. And yet the Internet remains. Think of the dead social media platforms: Friendster, Digg, Google Plus, Myspace and Vine. And yet social media is bigger than ever.

We all know who the big AI players are today: OpenAI, Google, Meta and Microsoft. But who’s to say they’re going to be the legacy AI players of tomorrow? It’s clear that AI is here to stay. But how big this bubble gets—and who survives—remains anyone’s guess.