Robert Cohen said that artificial intelligence debt is likely to reach bubble levels in credit markets during a forum in New York [1].
This warning suggests that the rapid accumulation of debt to fund AI infrastructure may eventually lead to a market correction. If a bubble forms, it could destabilize credit markets and impact the broader financial system as investors struggle to recoup their capital.
Speaking June 3, 2026, at the Bloomberg Global Credit Forum, Cohen, who serves as the director of global developed credit at DoubleLine, addressed the trajectory of AI-related borrowing [1, 2]. He said that the current cycle of investment mirrors previous historical patterns where massive capital injections into new technologies created unsustainable peaks [1, 2].
Cohen pointed to the history of railroads and the early internet as primary examples of this cycle. In those instances, heavy investment in infrastructure initially drove growth but eventually resulted in a bubble when the financial expectations exceeded the immediate economic reality [1, 2].
"Artificial intelligence debt will almost certainly reach bubble levels," Cohen said [1].
Despite the current optimism surrounding AI productivity and efficiency, the reliance on debt to scale these technologies creates a specific risk profile in the credit markets. While equity markets often absorb the first wave of volatility, the credit market deals with the actual repayment of the principal used to build the data centers, and hardware required for AI [1, 2].
Cohen's analysis emphasizes that while the technology itself may be transformative, the financial mechanisms used to fund it often follow a predictable path toward overextension [1, 2].
“"Artificial intelligence debt will almost certainly reach bubble levels."”
The warning highlights a shift in focus from AI's software capabilities to the physical and financial costs of its infrastructure. By comparing AI to the railroad and dot-com eras, the analysis suggests that the 'bubble' is not necessarily about the failure of the technology, but about the financial over-leveraging that typically precedes a market correction.
