New York, October 14 — Major U.S. and European banks, including JPMorgan Chase, Goldman Sachs, and Morgan Stanley, are seeking to offload portions of loans they underwrote last year to finance AI-focused data center projects, according to financial industry sources. The move comes as early signs of slowing demand for AI computing capacity emerge, particularly for specialized graphics processing units (GPUs) used in training large language models.
These financings, structured as senior secured term loans, were led by the aforementioned banks in late 2023 and early 2024 to fund data center expansions by companies such as CoreWeave, Lambda Labs, and Applied Digital. Loan sizes ranged from $500 million to over $2 billion, with maturities of five to seven years and interest rates tied to SOFR plus spreads averaging 300 to 500 basis points.
The total volume of AI-related data center debt underwritten by banks in 2023 and early 2024 is estimated at approximately $25 billion. Banks are now exploring secondary market sales of these assets through private placements to credit funds, insurance companies, and collateralized loan obligation (CLO) managers, seeking to free up balance sheet capacity and reduce concentration risk in the high-growth but volatile AI infrastructure sector.
The developments occur amid broader market volatility in technology stocks and rising long-term interest rates, which have increased the cost of carrying leveraged positions in speculative growth sectors. Historical precedent for similar bank-led debt sales includes the offloading of telecom and media loans after the dot-com boom and energy project financings during periods of commodity price downturns.
No banks have publicly disclosed specific sale amounts or buyers at this time. The reassessment among lenders of the risk profile of AI-driven capital expenditures follows a surge in spending following the public launch of generative AI tools in late 2022. However, these investments now face scrutiny over utilization rates, power consumption constraints, and the pace of enterprise AI adoption.
Looking ahead, investors and lenders will be closely watching developments in the AI sector, particularly as companies navigate the challenges of scaling their AI infrastructure while maintaining profitability and managing risk.
