KKR, a leading U.S. investment firm, warns about the growing pile of AI debt — SkimNews
Get the Finance newsletter
Daily finance — markets, central banks, M&A, the prints that move money. Free.
- KKR credit experts Christopher Sheldon and Tal Reback warned that AI infrastructure debt is highly concentrated, with power, chips, cooling, and financing often tied to the same small group of economic actors.
- 31 companies have issued over $500 billion in AI-related bonds as of August, with five issuers accounting for more than half the total, according to JP Morgan and KKR research.
- AI-related debt now totals around $600 billion, representing approximately 6% of the U.S. investment grade debt market, a figure KKR says could rise to 20% as capital expenditures swell.
- Stijn van Nieuwerburgh estimated that data center and AI infrastructure spending could reach $10 trillion from 2025 to 2032, surpassing past major U.S. industrial buildouts in scale relative to the economy.
- KKR analysts highlighted underappreciated risks including rapid depreciation of AI chips, multi-year data center construction delays, and power grid constraints that could limit AI deployment and revenue generation.
- Borrowers could struggle to repay debt if projected AI demand fails to materialize quickly enough, given the high cash burn and long lead times, making it difficult to 'cut back to health' once investments are made.
Why it matters: With AI infrastructure debt on track to make up 20% of the investment grade market, lenders and investors face outsized exposure to a handful of firms and physical bottlenecks—meaning a misstep in timing or demand could trigger widespread credit stress despite the sector’s long-term promise.
Ask SkimNews
