Are We in an AI Bubble? Here’s What 13 Economists Told Us

The last time this many smart people agreed something looked frothy, it was 2007 and they were talking about mortgage-backed securities. So when more than 75% of economists surveyed by NerdWallet say the AI investment boom looks like a bubble, it’s probably worth paying attention. But 2026 has brought new data, new wrinkles, and new reasons to question both sides of the argument. Here’s where things actually stand.

What Does “AI Bubble” Even Mean in 2026?

A bubble, in the simplest terms, is when asset prices inflate far beyond what the underlying fundamentals justify – driven by hype, speculation, and the fear of missing out. The tricky part: you can rarely confirm a bubble until after it pops.

The question of whether we’re in an AI bubble is one that 13 financial economists from universities across the U.S. weighed in on, and their answers were surprisingly lopsided. Ten out of thirteen said yes, they believe current AI investment levels constitute a bubble. Most gave their responses anonymously, given the sensitivity of making such a call publicly.

That doesn’t mean they’re right. But the reasoning they offered deserves a serious look, especially given how much has shifted in the past year.

The Bull Case: Why Some Economists Say This Isn’t a Bubble

Three of the thirteen economists pushed back on the bubble narrative, and their argument is straightforward: the biggest AI companies are actually making real money.

Here’s a snapshot of recent earnings from the three largest AI-linked stocks:

Those P/E ratios are elevated compared to historical norms (a “typical” ratio sits somewhere in the 20s), but they’re not absurd. For context, CVS Health has a P/E ratio above 200. All three tech giants have actually seen their P/E ratios decrease over recent quarters as earnings caught up with stock prices.

The anti-bubble camp essentially argues: this isn’t tulip mania. These companies are generating tens of billions in real revenue. The technology works. Customers are paying for it.

That’s a fair point. But it’s not the whole story.

The Bear Case: Why 10 Out of 13 Economists Said “Yes, It’s a Bubble”

The majority opinion was clear, and the economists cited several specific red flags. Here’s what concerned them most:

1. Venture Capital Is Pouring In at an Unsustainable Pace

The stock market gets most of the attention, but the private markets tell an even wilder story. Venture capital investment in AI startups surged 62% in a single year, according to Crunchbase data. Money is flowing into virtually any company that slaps “AI” onto its pitch deck.

If that sounds familiar, it should. The same pattern played out with dot-com startups in the late 1990s, crypto projects in 2021, and cannabis companies around 2018. The pattern is always the same: real technology attracts real interest, which attracts speculative money, which attracts dumb money.

2. The Circular Financing Problem Nobody Talks About

This is the one that should make you uncomfortable. Several economists flagged a pattern where Big Tech companies are essentially investing in each other’s AI ventures:

• Company A signs a multi-billion-dollar cloud computing deal with Company B
• Company B uses that revenue to invest in Company C’s AI hardware
• Company C reports record earnings, which drives up its stock price
• Investors see the rising stock prices and pour in more money

The question is how much of the reported AI revenue is organic demand from actual end users versus money being recycled between a handful of tech giants who are all flush with investor cash for the same reason.

If a significant chunk of that revenue is circular, the actual demand picture could be much weaker than the earnings reports suggest.

3. Valuations Outside the Top Three Are Genuinely Wild

Nvidia, Microsoft, and Alphabet might have defensible valuations, but look at the rest of the AI stock universe:

A P/E ratio of 400 means investors are paying $400 for every $1 of current earnings. That’s a bet on massive future growth that may or may not materialize.

4. The Dot-Com Parallel Is Uncomfortably Close

Here’s the thing that several economists emphasized: believing AI is transformative and believing we’re in a bubble are not contradictory positions. The internet really was transformative in the late 1990s. It genuinely did reshape the global economy. And investors still lost trillions when the dot-com bubble burst in 2000.

The technology being real doesn’t prevent the investment mania from being irrational. Those two things can coexist.

What’s Changed in 2026 That Makes This Question More Urgent

Several developments this year have sharpened the debate:

• AI spending commitments have ballooned. Major tech companies have collectively pledged hundreds of billions toward AI infrastructure, data centers, and chip development through 2028. If returns don’t match those outlays, the correction could be severe.
• Enterprise adoption has been slower than projected. While consumer-facing AI tools have exploded in popularity, many corporations are still in pilot phases rather than full deployment. The gap between AI hype and actual enterprise revenue generation is a growing concern.
• Interest rate uncertainty persists. Higher-for-longer interest rates make speculative growth bets riskier, since future earnings are worth less in present-value terms. This puts extra pressure on stocks trading at triple-digit P/E ratios.
• Regulatory scrutiny is increasing. Governments in the U.S., EU, and China are all moving toward tighter AI regulation, which could slow adoption timelines and add compliance costs that eat into margins.

How to Protect Your Portfolio If the Bubble Bursts

Here’s where things get practical. If you’re worried about AI concentration risk in your investments, the standard advice of “just buy index funds” might not be enough this time.

Why? Because the most popular index funds are heavily loaded with AI stocks. Nvidia alone makes up roughly 7% of the S&P 500 by market capitalization. The seven largest S&P 500 companies – all tech firms with major AI exposure – account for about a third of the entire index. The Nasdaq 100 is even more concentrated: Nvidia represents over 13% of it, and the top seven AI-linked stocks make up more than half.

That means a standard S&P 500 index fund isn’t as diversified as you might think.

Alternatives Worth Considering

Here are some options that reduce your AI concentration risk:

• Dow Jones Industrial Average ETFs: Because the Dow is price-weighted rather than market-cap-weighted, Nvidia accounts for less than 2.5% of it, and Big Tech overall is under 25%. That’s a meaningful difference from the S&P 500.
• Russell 2000 funds: This small-cap index is less than 14% tech. It’s heavily weighted toward industrial, healthcare, and financial companies. One caveat: small-cap stocks can be volatile, and a broad market panic triggered by an AI correction could still drag them down.
• Equal-weight S&P 500 ETFs: These hold all 500 companies in roughly equal proportions, which eliminates the mega-cap tech concentration problem. The trade-off is higher fees and potentially lower returns during bull markets.
• International ETFs: Funds like the Vanguard FTSE All-World Ex-US ETF (VEU) have less than 14% tech exposure. Geographic diversification adds a layer of protection against a U.S.-centric AI correction.

A quick disclaimer: none of this is personalized investment advice. Past performance doesn’t guarantee future results, and every investor’s situation is different. If you’re making significant portfolio changes, talking to a financial advisor is worth the time.

The Red Flags That Should Make You Nervous

Watch for these warning signs that a correction may be approaching:

• Earnings misses from major AI companies. If Nvidia or Microsoft report quarterly revenue below expectations, the market reaction could be swift and brutal.
• VC funding pullback. When venture capitalists start tightening their wallets on AI deals, it signals that the smart money is getting cautious.
• Rising insider selling. If executives at AI companies are selling their own stock at elevated rates, pay attention to what they’re doing rather than what they’re saying.
• Increased short interest. A spike in short positions on AI stocks suggests institutional investors are betting on a decline.

Frequently Asked Questions

Are we actually in an AI bubble right now?

The honest answer is that nobody can say with certainty until after the fact. However, 10 out of 13 economists surveyed by NerdWallet believe current AI investment levels constitute a bubble. Their concerns center on inflated venture capital spending, circular financing between tech giants, and historically high P/E ratios for many AI stocks. The technology itself may be genuinely valuable while the investment frenzy around it remains irrational.

What happens to my 401(k) if the AI bubble pops?

If your 401(k) is invested in standard target-date funds or S&P 500 index funds, you likely have significant exposure to AI stocks. A major correction could temporarily reduce your portfolio’s value, potentially by 20-30% based on historical bubble-burst patterns. The key word is “temporarily” – if you’re decades from retirement, riding out the downturn has historically been the better move compared to panic selling.

Should I sell all my AI stocks right now?

Timing the market is notoriously difficult, and selling everything based on bubble fears can backfire badly. Many economists who believe we’re in a bubble still acknowledge that the correction could be months or years away. A more measured approach might be rebalancing your portfolio to reduce AI concentration rather than making an all-or-nothing bet. Consider consulting a financial advisor before making major changes.

How is the AI bubble different from the dot-com bubble?

The biggest difference is that today’s leading AI companies are generating massive real revenue and profits, whereas many dot-com companies had little to no earnings. The similarity is the speculative frenzy: money flooding into anything AI-related regardless of fundamentals, sky-high valuations for unproven companies, and a widespread belief that “this time is different.” Both elements – real value and real speculation – can exist simultaneously.

What You Should Do This Week

Take 15 minutes to log into your brokerage or retirement account and check your actual AI exposure. Look at your top holdings. If Nvidia, Microsoft, Meta, and Alphabet collectively make up more than 25% of your portfolio, you’re making a concentrated bet on AI whether you intended to or not. That’s not necessarily wrong, but it should be a conscious choice rather than an accident of market-cap weighting. Knowledge of your own exposure is the first step toward making an informed decision about what, if anything, to change.