iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Assess AI lending concentration by looking through the portfolio—not just counting borrower names. Identify which borrowers’ earnings could be affected by AI, map shared customers, technologies and refinancing dependencies, measure exposures at several levels, and stress those common drivers alongside each borrower’s ability to repay. Then set limits and monitoring triggers suited to the portfolio’s risk appetite. There is no single standard AI-concentration definition or universal private-credit limit established by the sources cited here.
First, define what “AI lending concentration” means
The phrase can describe three different exposures. Choose the one you are measuring before calculating a percentage: they have different data needs and risk mechanisms.
| Exposure channel | What is concentrated | Primary question |
|---|---|---|
| AI disruption of borrowers | Loans to businesses whose revenue, pricing power or operating model could be affected by AI | Could AI adoption, substitution or changing customer demand weaken these borrowers’ cash flows? |
| Lending to AI-related businesses | Loans to companies developing or supplying AI products, services or infrastructure | Does the portfolio depend on the prospects, funding needs or valuations of AI-related businesses? |
| AI use by lenders | Use of AI models in underwriting, servicing or other credit decisions | Could model performance, governance or third-party dependence affect lending decisions or operations? |
The latest evidence discussed below concerns the first channel: business development company (BDC) loans to software firms facing uncertainty about generative AI. It does not establish the scale of AI exposure across every private-credit fund or answer questions about lenders’ own AI systems.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
What the available evidence says—and what it does not
BDC software lending is a signal, not a portfolio loss estimate
The Bank for International Settlements’ Bulletin 128, published 14 July 2026, reports around $115 billion in BDC lending to software firms—about a fifth of BDC lending and over 80% of their fast-growing technology portfolios. The bulletin says generative-AI-related revenue uncertainty had not affected these loans at the time of publication, and BDCs and their equity investors had not priced the software exposure differently. It also notes that credit spreads had recently narrowed, leaving less of a spread buffer against losses, and that some large BDCs shared borrower pools. Low leverage and secured lending may limit spillovers.
#1 Best Overall
These are sector observations, not estimates of AI-driven losses or a measure of any particular fund’s exposure. BDC findings should not be presented as representative of all private-credit vehicles. Use them as a reason to test for shared software and borrower exposures, not as a substitute for portfolio-level data.
Bank lending to private-credit vehicles is a separate risk channel
A Federal Reserve staff note published 23 May 2025 examines banks’ lending to private-credit vehicles, rather than the underlying loans in those vehicles. In its sample, commitment concentration was moderate; the note defines Herfindahl-Hirschman Index (HHI) values from 0 to 1, with higher values indicating less diversification. It also considers stress if vehicles draw unused bank lines. In a hypothetical full-draw scenario, the note estimates $36 billion in increased drawdowns, equal to about 2% of the Y-14 banks’ CET1 capital, with an aggregate CET1 ratio impact of roughly 2 basis points and an LCR impact of 1 percentage point. Those modeled results apply to the note’s scenario and sample, not to AI-driven portfolio losses.
The same note reports that leverage in its sample of 40 publicly traded BDCs rose from about 40% in 2017 to 53% in 2024. That is context for assessing financing sensitivity, not a measure of AI concentration. Keep both the populations and questions distinct when using these findings. Read the Federal Reserve staff note.
Recommended Free Tools
How to measure exposure across the portfolio
1. Set the perimeter and measurement basis
State exactly what is included: the fund, sleeves, co-investments, unfunded commitments, warehoused loans and relevant financing links. Record the measurement date and distinguish drawn balances from committed amounts and stressed exposures. Decide whether the analysis covers AI disruption of borrowers, lending to AI-related businesses, or both; do not combine them into one unexplained “AI” figure.
2. Build a look-through map
Aggregate connected borrowers and sponsors where appropriate, and classify each exposure by sector and software sub-sector, revenue source, customer and supplier dependencies, geography, maturity, seniority, covenant package, collateral and loan vehicle. Tag AI exposure using a documented rationale—for example, product substitutability, customer adoption or dependence on a particular technology. Record missing information as unknown; treating unknown exposure as zero can make apparent concentration misleading.
3. Calculate name, segment and overlap measures
Report largest-name and connected-group shares, sector and sub-sector weights, and top-N shares. Calculate HHI for clearly defined units—such as borrowers or sectors—and use a consistent exposure basis and methodology over time. An HHI summarizes how exposure is distributed across those units; it does not show whether borrowers share the same economic risks.
Add look-through overlap measures: borrowers appearing in multiple funds, common sponsors, shared end markets, and common AI-sensitive revenue drivers. A low share for each individual borrower can coexist with a large portfolio exposure to one shared driver, such as the same customer base or technology dependency. No single concentration statistic proves that a portfolio is safe.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →How to assess whether concentration can become credit loss
Sector exposure and credit quality are related but distinct. Compare AI-exposed borrowers with other portfolio segments across the factors that determine whether operating pressure can impair repayment or recovery.
- Cash generation: distinguish recurring revenue from discretionary or project-based revenue; examine customer concentration, margins and debt-service capacity.
- Balance-sheet resilience: assess leverage, liquidity runway, maturities and dependence on refinancing.
- Loan protection: review covenant headroom, collateral coverage, lien priority and realistic recovery prospects.
- Valuation and sponsor dependence: identify reliance on enterprise value, future growth assumptions or sponsor support to repay or refinance.
- Shared dependencies: compare borrowers’ customers, suppliers, technologies, end markets and financing sources, including dependencies that cross funds.
These comparisons help distinguish a high AI-sensitive exposure with strong repayment capacity from a smaller exposure whose borrowers have little room to absorb weaker revenue or refinancing conditions.
How to stress-test shared AI and credit risks
Do not assign precise AI-disruption probabilities unless they are supported by evidence. The cited sources do not supply AI-specific scenario probabilities. State assumptions transparently and use ranges where appropriate.
- Choose plausible operating shocks. Test product substitution, customer churn, pricing pressure, slower growth, margin declines and increased investment needs.
- Link operating pressure to financing conditions. Combine those shocks with higher financing costs, reduced refinancing availability, lower enterprise values and weaker collateral recoveries.
- Model common exposures together. Apply linked assumptions to borrowers that share customers, technologies, sponsors or end markets rather than treating every loan as independent.
- Trace the effects through loan protections. Estimate consequences for cash flow, covenant headroom, defaults, recovery and stressed loss, taking account of leverage, collateral and lien priority.
- Include liquidity and funding interactions. Where relevant, test cash needs and correlated draws on available credit lines alongside borrower-level stress.
- Compare results with limits and risk capacity. Show which names, segments or shared drivers breach thresholds and how the portfolio’s aggregate stressed exposure changes.
How to turn the assessment into controls
Translate the portfolio’s documented risk appetite into name, sector, sponsor and shared-factor limits or watch thresholds. Pair numeric limits with escalation triggers for deteriorating borrower data, rapid sector growth, limit breaches, covenant pressure, spread or valuation changes, and rising unknown exposure. Assign an owner, review frequency, independent challenge and a route for escalation to the investment committee or board.
Free tools Windows power users keep installed
One-click scans. No signup required.
Useful risk-management principles appear in U.S. interagency guidance, but their scope matters. The interagency commercial real estate concentration guidance supports sound segmentation, limits and sublimits, portfolio-level management, correlation analysis, timely management information, stress testing and contingency planning. Its scope is commercial real estate lending, so these are principles to apply by analogy—not rules that directly regulate private-credit funds. It cautions against splitting segments merely to obscure a concentration.
The interagency leveraged-lending guidance supports written, measurable underwriting standards; analysis of borrower sustainability; realistic downside scenarios; monitoring of covenants and collateral; and scrutiny of dependence on enterprise value. Use these principles to inform controls, while distinguishing guidance for its intended scope from requirements applicable to a particular fund.
What model-risk and consumer-credit rules do—and do not—cover
The OCC’s 2026 revised model-risk guidance discusses model development and use, validation and monitoring, governance and controls, and vendor or third-party products. It is non-prescriptive, expressly excludes generative and agentic AI models from its scope, and is most relevant to banking organizations. It is not an AI concentration rule for private-credit portfolios.
Consumer-credit obligations are also a different issue. The CFPB’s 19 September 2023 guidance says lenders using complex algorithms must give accurate, specific reasons for adverse actions; a broad checklist item may not suffice if it does not reflect the actual reason. That is a consumer-credit disclosure point, not a private-credit portfolio concentration standard.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
