BIS warns AI bust could ripple into credit; hyperscalers shift from cash-flow to debt financing

The Bank for International Settlements warned that an artificial-intelligence bust, inflation, and fiscal stress rank among the most alarming threats to global prosperity. The BIS noted that AI infrastructure spending—already reaching $400+ billion annually from the five largest US tech firms (Amazon, Alphabet, Microsoft, Meta, Oracle)—has become the primary driver of capex growth and GDP expansion. But the financing model is shifting: hyperscalers are increasingly turning to debt issuance rather than cash flows to fund multi-year data center buildouts.

AI hyperscalers roughly doubled capex in 2025–2026 and are financing this surge largely through corporate bonds, with gross issuance exceeding $100 billion in 2025. Most issuance was long-term (5+ years), locking in funding for multi-year infrastructure programs. However, credit default swap (CDS) spreads on AI hyperscalers have risen, signaling investor concern about debt sustainability. Morgan Stanley estimates that nearly $3 trillion in AI-related infrastructure investment will flow through the global economy by 2028, with over 80% of spending still ahead.

The structural risk is that if AI monetization lags capex returns, or if a funding shock occurs (e.g., private credit lenders pulling back), the debt burden could trigger cascading defaults and credit market disruption. The BIS identified private credit as a particular vulnerability—lenders have extended enormous sums to untested borrowers using unconventional collateral. For practitioners, the takeaway is clear: infrastructure builders who cannot achieve positive unit economics or demonstrate path to profitability within 12–18 months face refinancing risk.