JPMorgan has issued a stark warning that AI companies can no longer rely on their label alone to secure financing, as lenders increasingly demand concrete evidence of financial viability. According to JPMorgan global investment strategist Justin Biemann, companies must now demonstrate they can finance investments, service debt, and generate cash flow before securing funding. This represents a fundamental shift from the earlier period when AI companies could access capital based solely on their AI-related business model. The warning comes as credit-default-swap trading tied to major AI companies surged in September, with the cost of insuring some debt-dependent firms reaching levels signaling heightened default risk.
The financial stress is evident in unprecedented credit default swap (CDS) spreads that have reached record levels. As reported by JPMorgan, one financing-dependent AI infrastructure company has taken on tens of billions of dollars in debt and other obligations, with its five-year CDS jumping to 1,100 basis points. This means insuring $10 million of its debt would cost more than $1 million a year. Pricing at these levels implies near or above a 50% probability of default over five years, up from below 30% a year ago, according to Biemann. The scramble for CDS protection is evident in trading volumes, with weekly trading in CDS linked to bonds issued by Oracle, SpaceX, Alphabet, Amazon, and Meta showing significant activity. Large AI leaders with strong cash flow pay 50 to 120 basis points for five-year protection, while some established infrastructure borrowers have climbed toward 300 basis points.
The AI financing crisis is reflected in the tech-heavy Nasdaq sliding 6.7% to 29,263 from 31,360 since October 6, with AI-linked firms such as Nvidia and Oracle falling more than the index. Elon Musk issued a similar warning on Sunday, saying debt funding is drying up for AI neoclouds, firms whose revenue depends almost entirely on data centers. The stress is particularly concentrated among businesses reliant on continued borrowing, though JPMorgan warns that a deeper AI downturn could spill into other risk assets, including Bitcoin. Biemann noted that AI has been one of the most potent narratives driving risk-taking across financial markets, with Bitcoin and Nvidia both bottoming within weeks of ChatGPT's debut in late 2022 and rallying significantly since.
Several companies are reported to have significant debts related to AI, including Amazon, which is estimated to owe ₹4,400 crore (53 billion US dollars) from the bond market during 2025-2026. The spread between five-year CDS yields for CoreWeave, Oracle, and SoftBank continues to widen, with CoreWeave's bond yields exceeding 15 percent and the company's stock falling by 24 percent two weeks ago. Oracle is also facing delays or potential failures in the construction of several data centers due to licensing issues. In September 2026, SoftBank raised a record ₹9,000 crore (11 billion US dollars) in debt, with interest rates nearing 10 percent after making significant investments in OpenAI. The accumulation of massive debt is changing the financial rules for companies, involving complex networks that include major Wall Street banks, private credit providers, public bond markets, and various intricate financing schemes.
The AI boom carries greater risks than previously anticipated, with the market starting to realize a potential mismatch between money committed and future income-earning ability. As reported by The New York Times, there is a collective awareness regarding AI developments, with concerns triggering a sell-off in AI company stocks, widened corporate bond yield spreads, and warnings on credit risk indicators. These concerns have driven a surge in insurance from various companies to protect against AI investment default risk. However, some investors are finding silver linings in the current situation. John Lloyd, Head of Global Multi-Sector and Corporate Credit at Janus Henderson, noted that the slowdown in AI costs makes existing AI projects and infrastructure more valuable. The risks associated with AI debt financing are widespread across the global financial ecosystem, with both institutional and retail investors holding AI loans and bonds, though risks for each are situated on different ends of the risk spectrum. The latest credit default risk projections suggest that OpenAI bankruptcy probability could reach 17.5% by 2029, indicating sustained financial pressure on AI infrastructure companies.
