AI Private Credit Risks
EXECUTIVE SUMMARY
AI-driven disruption poses material risks to private credit portfolios, particularly those with heavy exposure to software companies, which comprise ~21% of the asset class and up to 40% including broader tech services. While defaults remain historically low, mispriced pre-2024 loans and eroding SaaS margins could elevate loss rates over 3-5 years, with 25-35% of portfolios facing elevated AI risk. Investors should prioritize managers with disciplined underwriting and restructuring expertise amid rising dispersion.
MARKET CONTEXT
Private credit has ballooned to over $1.7 trillion in assets under management as of late 2025, fueled by demand for higher yields amid compressed public market spreads, but faces intensifying scrutiny from AI-related headwinds. Software exposure has grown steadily over 15 years, reaching ~21% in private credit—higher than in leveraged loans—driven by SaaS models' appeal: recurring revenues, ~96% private-held firms, and fat margins that lured aggressive lending. Including business services/tech, exposure hits ~40%, dwarfing prior retail sector vulnerabilities that pressured credits in 2016-2019; recent BDC underperformance reflects this, with laggards showing subpar risk controls and software susceptibility. Opaque, unrated loans amplify uncertainty, as pre-2024 originations overlooked AI threats amid refinancing booms and direct lending surges that compressed spreads.
KEY DEVELOPMENTS
- On February 4, 2026, Ares Capital CEO Kort Schnabel acknowledged AI's threat to single-function software firms in content/analysis tools, yet reported no portfolio weakness, signaling lenders' heightened disclosures amid investor probes.
- A mid-2025 report pegged 25-35% of private credit portfolios at elevated AI disruption risk, with public equity volatility spilling into private markets via cash-flow durability concerns over 5-year loan horizons.
- Software's relative heft—bigger than retail—was flagged as a more consequential stressor than 2016-2019 woes, with non-uniform disruption expected to lift defaults gradually over 3-5 years, alongside refinancing pressures in a tightening liquidity backdrop.
INVESTMENT IMPLICATIONS
Portfolio positioning warrants reduced beta to software-heavy strategies; favor managers with <15% sector exposure, strong covenants, and proven recovery track records, targeting 9-11% net yields while stress-testing for 200-300bps spread decompression. Allocate 5-10% to private credit for illiquidity premia, but tilt toward diversified or "mission-critical" SaaS lenders embracing AI for growth—potentially boosting profitability—over legacy single-function plays; secondary markets and structured tools offer tactical liquidity amid redemptions. This setup exploits dispersion: top-quartile managers could deliver 150-200bps excess returns versus averages as weaker credits surface cyclically.
RISKS TO MONITOR
- AI Disruption Intensity: Non-uniform but potent margin erosion in ~30% of SaaS models could impair cash flows, with pre-2024 loans underpricing risks and recovery values at stake over 5 years; monitor for clustered defaults rising from <2% baselines.
- Refinancing and Liquidity Squeeze: Tightening conditions expose rollover risks in software, amplified by opacity and redemption pressures; dividend cuts in vulnerable BDCs signal broader yield normalization.
OUR VIEW
While headlines amplify AI panic, the real edge lies in asymmetry: disciplined managers will harvest outsized recoveries from restructurings, turning 25-35% portfolio stress into alpha as defaults creep to 4-6% selectively—not systemically—rewarding pre-2024 mispricing. Contrarian tilt: underweight pure software BDCs, overweight multi-sector originators with AI-adaptive borrowers; private credit's illiquidity buffer endures, but only for those stress-testing now.
