Ray Dalio says AI is ‘a classic bubble’ nearing its burst — debt, not the models, sets the clock

Speaking at the Forbes Global CEO Conference in Singapore on Wednesday, the Bridgewater founder argued that the AI buildout is increasingly financed with debt, that rising bond yields are raising the cost of data centers and chips, and that “we’re close to” the bursting point — even as the S&P 500 and Nasdaq 100 hit record highs.

Ray Dalio doesn’t think the AI boom ends when the models stop improving. It ends when the financing does. Speaking at the Forbes Global CEO Conference in Singapore on Wednesday, the Bridgewater founder told a room of global executives that artificial intelligence is “a classic bubble” nearing its bursting point — with debt, not demand, as the trigger.

“We are in the part of the cycle that is before that but approaching that,” Dalio said, per an IANS dispatch from the conference. “I think we are close to that.” The argument: technology companies have been spending hundreds of billions of dollars on AI, and an increasing share of that spending is being financed with debt. Meanwhile, rising bond yields around the world are pushing up the cost of funding AI infrastructure — data centers, chips, power. Debt gets more expensive exactly when the industry needs the most of it.

Debt is the fuse

Dalio’s core mechanism is financial, not technological. A large amount of debt is being raised to fund AI investment, and market gains have become concentrated in a small handful of large tech firms. When the pressure builds, he argued, the wealth locked up in those valuations is harder to realize than it looks: “Everybody says ‘I’m worth a billion dollars’ but OK, try to spend that,” he said. People would need to sell assets to turn wealth into cash, and that selling is what pressures bubbles into bursting.

He flagged other potential triggers too, including wealth taxes and the rush to convert unrealized gains into cash. The framing matches a warning he issued in June, when he wrote on X that the United States is entering a particularly risky stretch between the 2026 midterm elections and the 2028 presidential election, driven by widening fiscal deficits, rising debt and falling demand for US government bonds: “I believe we are currently on the brink.”

The productivity question

Dalio wasn’t the only skeptic on stage. He appeared alongside executives from Franklin Templeton, Temasek Holdings and Bangkok Bank, and Franklin Templeton CEO Jenny Johnson said AI hasn’t yet produced real productivity gains for U.S. companies — a pointed rejoinder to the thesis that enterprise AI spending pays for itself, per CoinCentral’s writeup of the session. Markets, meanwhile, kept climbing: the S&P 500 and Nasdaq 100 hit record highs this week even as Dalio issued the warning, buoyed by optimism over technology earnings.

His disclosed hedge is characteristically Dalio: spread bets across 10 to 15 uncorrelated streams — and short debt.

Why it matters for AI

The bubble debate is no longer about whether the technology works — it’s about how the buildout is financed. If AI infrastructure is increasingly debt-funded and concentrated in a few names, then the industry’s fate is tied to bond markets and interest rates as much as to model benchmarks. Dalio’s claim is really about sequence: the models can keep improving while the financing structure deteriorates, and the burst comes from the latter, not the former. For anyone tracking the AI investment cycle, Wednesday’s session was the highest-profile airing yet of the case that the constraint isn’t chips or talent — it’s credit.

Sources: IANS via hi INDiA, The Cryptonomist (citing CoinCentral).