AI bubble burst in 2026?

Prediction markets put the probability at 22%: AI bubble burst in 2026. Currently, markets see this as unlikely (22% YES). Three ways the AI boom may fizzle — and what happens to stocks next - MarketWatch.

What’s Happening

As of May 2026, prediction market participants are assigning a 22% probability to the event of an "AI bubble burst" occurring this year, with a strong 78% majority betting against such a collapse. This cautious optimism follows a turbulent week of mixed signals from industry leaders. On May 17, 2026, Citadel CEO Ken Griffin, a former prominent AI skeptic who had dismissed the technology as "garbage" at Davos in January, publicly reversed his stance. Speaking at Stanford Business School, Griffin admitted AI had become "profoundly more powerful," calling it "real" for the first time — a shift that bolstered confidence in the sector's fundamentals. However, a MarketWatch report from May 18, 2026 outlined three scenarios for how the AI boom could fizzle, including a "tech-investment recession" scenario where a correction in U.S. tech investments could trigger a 45% drawdown in tech hardware stocks and an 18.6% drop in the S&P 500, underscoring persistent downside risks. [MarketWatch, May 18]

The debate over whether an "ai bubble burst in" 2026 is imminent intensified on May 20, 2026, when Amazon founder Jeff Bezos directly addressed the concern. In an interview with CNBC, Bezos brushed off fears of a looming AI bubble, stating, "You shouldn't worry about it," and argued that massive capital expenditures by hyperscalers — projected at roughly $660 billion in 2026 — would ultimately push the technology forward. His comments came during the same CNBC "Opening Bell" segment, where he also discussed Blue Origin's potential outside investment and the future of space exploration. Bezos's reassurance provided a counterweight to bearish forecasts, but the market's 22% YES probability suggests a minority of participants still see a significant chance of a correction, particularly given the MarketWatch analysis that even a mild "correcting the excess" scenario could see a 4.5% drop in U.S. tech investments, or roughly $66 billion. [CNBC, May 20]

Looking ahead, the key question is whether the current consensus against an "ai bubble burst in" 2026 will hold as new data emerges. A Forbes report from May 18, 2026 highlighted a growing concern that AI itself could distort market signals, noting that "synthetic polling companies are now a dime a dozen" and that AI-generated data may make traditional research "less accurate and more misleading." This raises the possibility that prediction market probabilities could shift rapidly if actual economic indicators — such as hyperscaler capex reports or tech earnings — diverge from AI-influenced forecasts. Meanwhile, Griffin's dramatic about-face and Bezos's public confidence have provided a floor for sentiment, but the MarketWatch scenarios serve as a reminder that even a small correction in the overheated AI sector could cascade into broader market losses. The next major test will likely come with quarterly earnings reports from major tech firms later this year. [Forbes, May 18]

Traded on Polymarket — $2.3M Volume

Active market on Polymarket with $2.3M in total volume. Sufficient liquidity for most position sizes. Currently priced at 23c YES.