FSB Warns Private Credit Exposure to AI Boom Creates Systemic Financial Risk
Summary
- • AI companies now account for over a third of private credit deals in 2025, up from 17% over the prior five years
- • FSB warns a sharp correction in AI asset valuations could trigger sizeable credit losses to private credit investors
- • Electricity supply shortfalls named as a concrete risk trigger that could delay or cancel datacenter projects
- • Traditional banks increasingly entangled with private credit through lending and partnerships, amplifying contagion risk
Details
FSB releases formal report warning private credit sector is dangerously concentrated in AI financing
The Financial Stability Board monitors financial stability across 24 countries including all major central banks. A formal FSB report signals that AI financing risk has crossed from sector-level concern to a potential systemic issue warranting regulatory attention.
AI companies' share of private credit deals jumped from 17% (five-year average) to 33%+ in 2025
The doubling of AI's share within a single year indicates rapid concentration of credit risk in one sector. Private credit markets are already less liquid and less transparent than public debt markets, making this concentration harder to unwind in a stress scenario.
Sharp correction in AI asset valuations identified as primary credit-loss trigger for private credit investors
The FSB's language — 'sizeable credit losses' — reflects concern that current valuations have run ahead of fundamentals. Private credit investors cannot easily sell positions the way bond or equity holders can, meaning losses would be realized slowly and opaquely.
Electricity supply shortfalls named as a concrete trigger that could halt datacenter projects and impair loans
AI datacenter construction is power-constrained in many markets. The FSB's inclusion of energy supply as a named risk factor signals that physical infrastructure bottlenecks are now considered a financial stability variable, not just an engineering challenge.
Datacenter oversupply relative to demand is a second scenario that could compress AI company valuations and loan quality
If buildout accelerates faster than AI adoption generates revenue, the assets securing private credit loans lose value simultaneously with borrower cash flows. This dual-impairment dynamic is what makes infrastructure overbuild cycles historically severe.
Private credit borrowers structurally carry lower credit scores and higher debt loads than traditional bank borrowers
Private credit emerged partly to serve companies that cannot access investment-grade bond markets on favorable terms. The FSB is flagging that this inherent risk profile is now scaled to a size where losses could be system-relevant.
Traditional banks deeply embedded in private credit through direct lending, portfolio financing, and asset-manager partnerships
JP Morgan, Barclays, UBS, and Jefferies all suffered material losses from the Tricolor and First Brands collapses. Banks are not observers of private credit risk — they are participants with often incomplete borrower information.
Tricolor and First Brands, two private credit-backed US auto companies, collapsed with fraud allegations, burning major bank creditors
JP Morgan and Barclays took losses on Tricolor; UBS and Jefferies disclosed significant exposures. The FSB cites these cases to demonstrate that information gaps in private credit — lenders 'may have only partial information about borrowers' — translate directly into unexpected bank-level losses.
Some private credit funds forced to cap client withdrawals after a surge in redemption requests
Withdrawal caps are a liquidity stress signal. Private credit funds hold illiquid assets; a mismatch between investor redemption rights and asset liquidity can force distressed asset sales or gates, further undermining confidence in the sector.
Policy = regulatory/governmental action, Stat = quantitative data point, Insight = analytical conclusion or risk assessment, Infrastructure = physical/energy systems, Market Impact = valuation or demand dynamics, Industry Update = structural business development, Legal = litigation or fraud
What This Means
The FSB's intervention marks a formal escalation of AI financing risk from a sector concern to a systemic financial stability issue — the kind of designation that typically precedes regulatory action or mandatory disclosure requirements. For AI infrastructure investors and datacenter developers, the report introduces two compounding risk vectors that have not been widely priced: energy supply constraints that could physically impair projects, and the opacity of private credit structures that leaves even the banks financing them with incomplete exposure data. If valuations correct or power constraints materialize, the losses would ripple through institutions that most AI industry participants would not have considered directly exposed.