Even AI’s most enthusiastic backers are beginning to confront an uncomfortable truth. Signs of an economic bubble are flashing, and the stakes extend beyond Silicon Valley. If the AI boom falters, will the broader U.S. economy stumble with it? That question is no longer theoretical. It’s a concern voiced by investors, economists, CIOs and business leaders across the country.
Jeremy Kranz, founder and managing partner at Sentinel Global, said that it’s “hard to be definitive about AI holding up the entire economy,” but the sector’s circular economy — which refers to AI companies and data center companies investing in each other — has a “trickle-down economic impact” on supporting businesses. These include general contractors, housing builders and providers for people running these data centers, employees and the retailers and restaurants they buy from, and so on.
“When you’re talking about potentially $1 trillion of spend happening in the economy around one particular theme and sector — and that is AI — recognizing the trickle-down economics does support the belief that, in fact, we have the entire U.S. economy propped up,” Kranz said.
However, the timing of a bubble burst is still under debate.
According to Christopher Hodge, chief economist of the U.S. at Natixis CIB Americas, “this is likely not a risk for 2026” and “while at some point, the wind will come out of the sails of AI and optimism may fade, that is not likely a near-term story.” The reason? Hodge said, “Hyperscalers are in an arms race, and CapEx intentions for 2026 are still sky high and fueled in part by favorable tax changes from the One Big Beautiful Bill.”
CIOs are struggling to see a way forward in either scenario: a relatively contained AI bust, or an AI bust plus a larger economic hit. However, all is not likely lost if CIOs reshape their strategies and budgets now, while they still have time to be proactive.
Christopher Hodge, chief economist of the U.S., Natixis CIB Americas
The forces at work
Evidence of strain is mounting on multiple fronts. Some warn the first rupture is tearing through the labor market already, where massive and sudden layoffs threaten to erode consumer confidence and spending. Employers had cut 1.171 million jobs in 2025 by December, a 54% jump compared with 2024, according to a Reuters report. Relatedly, more than 7 million Americans are unemployed, the highest figure since 2017 when excluding the pandemic years. And while the headline unemployment rate remains relatively modest at 4.4%, it sits nearly a full percentage point above recent years. That’s a clear sign that finding work is becoming much harder for both the laid-off experienced workers seeking new work and entry-level workers trying to enter the workforce.
Investment supply vs. demand
Others see the potential crash forming inside the AI sector itself. Economist, London Business School lecturer and author Rebecca Homkes said the debate is not over the failings of the technology, despite the well-known frustrations with GenAI hallucinations and errors, but that instead “we are debating the hype cycle of the current belief of the AI hyperscalers.”
“The investment in supply versus tangible demand is where the nuance lies, and the debate that matters is the speed and timing of adoption by organizations,” Homkes added.
AI adoption rates are rising overall, but the sheen is wearing off for large enterprises where the big money lives. The Census Bureau’s Business Trends and Outlook Survey shows that AI use reached 10% of U.S. businesses in September, an increase from 3.7% a year earlier. But AI adoption among large enterprises slowed noticeably over the summer, as many production deployments failed to generate meaningful ROI.
“Announcements by credible players [that] they are pulling back on AI investments will shake this market,” Homkes said. “The current challenge we have: For every report showing increase in AI adoption and tangible gains, we have another one showing the lack of ROI.”
Indeed, adoption hesitation is starting to show up in earnings reports — even from the hyperscalers funding the wave of new data center construction across the country. Case in point: Microsoft’s stock recently dipped 3% after reports indicated the company has yet to see revenue growth catch up with its massive AI investments. The possibility of a broader domino effect, dragging down not just the AI sector but adjacent industries, has sparked concerns about stalled investment cycles, falling valuations and billions of dollars’ worth of AI data centers sitting underutilized.
AI and GDP
Alfonso Berumen, practitioner of decision sciences at Pepperdine University’s Graziadio Business School, said that while AI investment is boosting productivity and capital spending, “it is not the sole force holding up the U.S. economy.” However, “growth is a different story,” he said, adding that recent estimates suggest AI-related investment accounted for more than two-thirds of the 1.6% annualized GDP growth in the first half of the year.
“This indicates that while AI is not the foundation of the economy, it is disproportionately responsible for the incremental growth we are currently seeing. If AI investment slows, headline GDP could weaken quickly because other sectors are contributing far less to marginal growth,” Berumen added.
From the high view, these combined forces seem to teeter toward a likelihood that AI may come crashing down on top of an already uncertain U.S. economy, one plagued by shifting tariffs, growing inflation, an increasing number of business closures, and rising unemployment. But appearances can be deceiving, requiring a deeper dive into what may be happening.
Bubble repercussions
An economic bubble is a period when current asset prices dramatically exceed their intrinsic valuation, but there is no formal criteria with which to calculate a bubble. Dan Buckley, a chief analyst at DayTrading.com, said he has assessed whether AI is in a bubble “by looking at seven metrics,” including whether:
-
Prices are high relative to traditional measures.
-
Bullish sentiment is broad.
-
Purchases of assets are commonly done with high leverage
After examining these metrics, Buckley concluded that “pockets of the AI sector are certainly in bubble territory.” However, the U.S. economy “as a whole isn’t necessarily doomed,” he said.
Buckley’s assessment falls notably short of being a full-throated reassurance and it is certainly not in alignment with hyperscaler AI hype. But what lies beneath calls like his is a starker reality than may be readily apparent.
Paul Ferrara, a chartered investment manager at Avenue Investment Management, said “a pullback in AI may still spill over to supply chains, data centers building and credit markets dependent on technological growth.” He suggested that CIOs who want to “avoid a painful snapback” may want to “focus on durable gains instead of rapid [AI] expansion.”
Rich Pleeth, CEO and co-founder, Finmile
CIO budget strategies
Forewarned is forearmed and experts are advising CIOs to take action now to prepare for a harsh wake-up call to budget realities, no matter how the pending fate of the AI bubble and overall U.S. economy works out.
Rich Pleeth, CEO and co-founder of Finmile, a transportation logistics firm that offers AI-based delivery and route optimization, said that if the AI momentum slows, “the softness in other sectors will show up fast.” The risk is not that AI disappears, he said, but that AI is no longer “a blank check.”
“For CIOs and companies, the safest move is — and has been — to prioritize AI, as the markets presume that delivers cost reductions and operational efficiency. The projects that survive a slowdown will be the ones tied directly to unit economics, not the vanity experiments,” Pleeth said.
All told, the consensus is that a total freakout and AI cutback is not likely warranted.
“Bubbles burst, but industrial revolutions don’t,” said Jason Wild, a former executive at Microsoft, Salesforce and IBM, and co-author of “Genius at Scale” with Harvard Business School professor Linda A. Hill.
Wild said he knows that a correction is coming, given actions such as OpenAI “burning $5B annually while spending more than $2 for every $1 earned.” But like the earlier dot-com bust, “this shakeout will accelerate AI’s transformation, not end it.”
Wild said he sees the situation as a paradox, with most CIOs retreating to cost-cutting in anticipation of a bubble burst. But he predicted others will chart a braver course, while their competitors hunker down.
“The boldest will architect system-level change through frugal experimentation, prepare to acquire strategic assets at a fraction of peak valuations, and co-create the future where they can be world-class,” Wild said.
