Oracle may be facing a bank-debt “problem” due to its deal with the maker of ChatGPT, OpenAI. This comes as American banks struggle to absorb the scale of financing tied to Oracle’s data centre expansion plans. According to a recent report by The Wall Street Journal, the issue stems from a $300 billion agreement between Oracle and the Sam Altman-led OpenAI, which requires massive borrowing to fund data centres in Texas and Wisconsin. JPMorgan Chase and other US banks have reportedly found it difficult to distribute these loans due to exposure limits linked to a single borrower.Last month, Oracle’s leadership sent an email at 6AM to inform its employees about its plans to cut tens of thousands of jobs. Reports suggest that around 10,000 employees have already been affected, based on internal tracking of staff activity. Estimates from TD Cowen indicate the total number of layoffs could reach between 20,000 and 30,000, which would account for a notable share of Oracle’s workforce of about 162,000 employees.
What pressure banks are facing from large Oracle-linked loans
The WSJ report cited people familiar with the matter to claim that lenders spent months trying to spread the risk of billions of dollars in loans tied to Oracle-backed data centre projects. Many financial institutions have internal limits on how much exposure they can take to one company, and the scale of Oracle-related borrowing has pushed those limits.As a result, some bank balance sheets became constrained, affecting their ability to finance additional projects linked to Oracle and OpenAI. In one example, lenders were hesitant to back an expansion of a data centre complex in Abilene, Texas, if Oracle was the tenant. The company that built the data centre, Crusoe, eventually leased the facility to Microsoft.The financing challenges highlight broader risks in the growing data centre sector, where demand for AI infrastructure is increasing while access to capital remains uneven.The report claims that Oracle has attempted to address concerns by outlining plans to raise around $50 billion through stock and bond offerings to support its 2026 funding needs. The company said, “We are proud of the rapid progress that’s been made both in financing and developing our data centers,” adding that its partners have diversified funding sources.However, analysts at Morgan Stanley estimate that Oracle could require an additional $100 billion or more through 2027 and early 2028. “We’ve pondered how [Oracle’s] considerable funding needs over the next three years may test the depths of different fixed-income markets,” analysts noted.Compared to other major technology firms, Oracle is seen as having a weaker financial position, with higher debt levels and lower credit ratings. This has contributed to caution among lenders, particularly as the company increases its reliance on large-scale AI infrastructure projects.However, the financing strain is not limited to Oracle. The broader AI industry relies heavily on external funding to build data centres that provide the computing power required for AI systems.Analysts estimate that major technology companies may only be able to cover about half of the projected $3 trillion in AI-related spending through 2028 using their own cash flows. The remainder is expected to come from banks, bonds, and private credit markets.While companies such as Google, Microsoft, and Meta continue to attract strong lender support, Oracle’s situation shows that not all players have the same level of access to capital.The challenges also suggest that financing delays could slow the pace of data centre construction, potentially affecting the expansion plans of AI companies that rely on these facilities to scale their services.
