Ray Dalio, the founder of Bridgewater Associates, said the artificial intelligence market is in a bubble that will eventually burst [1].
This warning suggests that the risk to AI investments is not based on the utility of the technology itself, but on the financial mechanics of liquidity. If a large number of investors suddenly require cash, the resulting forced sales of overvalued assets could trigger a market collapse.
Dalio discussed these views on June 3, 2026 [1]. He noted that the current market trajectory follows a historical pattern seen with other major technological shifts. "All great technology changes produce bubbles, and the AI market is no exception," Dalio said [2].
The core of the risk lies in the distinction between wealth and money. Dalio said the bubble remains intact as long as assets are held, but the situation changes when those assets must be liquidated to meet cash needs. "When wealth is converted into money, the AI bubble will burst," Dalio said [1].
This perspective shifts the focus away from whether AI can deliver on its promises. Instead, it highlights the danger of inflated asset prices that cannot be sustained during a liquidity crunch. Dalio said the technology could be successful while investors still suffer losses due to timing and valuation [5].
"The real danger isn’t that the technology fails, but that investors will need cash and be forced to sell inflated AI assets," Dalio said [3]. He said the systemic risk is tied to the pressure of selling these assets into a market that may not have the depth to absorb them without a price crash.
“"When wealth is converted into money, the AI bubble will burst."”
Dalio's analysis distinguishes between technological viability and financial stability. By framing the AI bubble as a liquidity issue, he suggests that even if AI fundamentally transforms the global economy, the financial instruments used to bet on that transformation can still fail. This implies that the 'bursting' of the bubble is a function of investor behavior and cash flow requirements rather than a failure of the software or hardware.




