Ray Dalio has warned that the AI investment boom could break when investors need real cash, not when the technology itself disappoints.
Summary
- Ray Dalio warned that the AI bubble could burst when investors are forced to turn wealth into cash.
- Bridgewater estimated major technology companies could spend $650 billion on AI infrastructure in 2026 alone.
- Dalio linked the risk to debt pressure, tax debates, fund redemptions, and potential disruptions to Taiwan chip exports.
Bloomberg reported that the founder of Bridgewater Associates issued the warning during a television interview, in which he argued that bubbles often end when holders of valuable assets are forced to convert paper gains into spendable money. Dalio said the danger posed by artificial intelligence lies less in the quality of the technology and more in how financial markets fund fast-growing wealth.
According to Dalio, investors often confuse wealth with money. A private company may gain a billion-dollar valuation after raising a much smaller amount of capital, yet that valuation cannot be spent unless shares are sold. In his view, the pressure begins when many holders simultaneously try to convert that wealth into cash.
Dalio flags cash risk behind AI boom
During the Bloomberg interview, Dalio said major technology changes usually create bubbles because investors struggle to value the opportunity correctly. He said companies often spend heavily to win market share, even before the final winners are clear.
Bridgewater estimated that Alphabet, Amazon, Meta, and Microsoft could spend about $650 billion on AI infrastructure in 2026. The firm compared that with roughly $410 billion in 2025, showing how much large technology companies may commit to data centers, chips, and related systems.
Dalio tied that investment scale to a fragile market setup. He said paper wealth can rise much faster than the money supply, leaving investors exposed when debt payments, tax bills, or fund withdrawals force them to sell.
Government debt adds to the strain
In the same interview, Dalio connected the AI bubble risk to pressure on the U.S. government balance sheet. He said the United States spends about $7 trillion while collecting about $5 trillion in revenue, leaving the government dependent on additional borrowing.
Dalio said that heavy borrowing can strain the bond market, especially when long-term rates rise against short-term rates. He has made similar arguments in past warnings about debt, inflation, and the global monetary system.
His bubble indicators, according to Dalio, now sit near levels seen before the 2000 dot-com crash and the 1929 market collapse. He did not frame that as a reason for immediate panic, but he said investors should prepare for weaker returns.
Election timing could raise market tension
Dalio also pointed to a sensitive political window after the midterm elections and before the next presidential vote. He said tax debates during that period could pressure wealthy investors if policymakers push for changes that require asset sales.
At the same time, Dalio said that outside shocks could accelerate a downturn. He warned that any halt in chip exports from Taiwan would hit AI stocks hard because many AI companies depend on an advanced semiconductor supply.
Dalio’s argument also reaches crypto markets, where investors often hold assets whose value depends on liquidity. Bloomberg noted that Dalio still prefers Bitcoin, which he has described as digital gold, over cash.
