The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Assess AI lending concentration on a look-through basis: identify which borrowers could be affected by AI, measure both individual and shared-driver exposures, and test how correlated business pressure could translate into defaults and losses. A list of borrower names or a single sector percentage is not enough. There is no standard AI-concentration definition or universal limit for private-credit funds, so the assessment should state its scope, assumptions and risk thresholds clearly.
First define what “AI concentration” means
The phrase can describe three different exposures. They should not be combined into one figure without explaining what is being measured:
As an Amazon Associate I earn from qualifying purchases.
| Exposure channel | What it measures | Why it matters |
|---|---|---|
| Borrowers vulnerable to AI disruption | Loans to businesses whose revenue, pricing power, costs or operating model could change as customers adopt AI or AI substitutes for existing products. | This is the channel addressed by the recent evidence on BDC lending to software firms. |
| Lending to AI-related businesses | Loans to companies building, supplying or operating AI products and infrastructure. | This is a different sector exposure; the evidence cited here does not quantify it. |
| Lenders’ use of AI | AI models used in underwriting, servicing or other lending decisions. | This is a model governance and operational-risk question, not the same as concentration in a loan portfolio. |
For a portfolio assessment, document the perimeter before calculating anything: which fund and sleeves are included, whether co-investments, warehoused loans and unfunded commitments count, and whether connected financing links are in scope. Record the measurement date and distinguish drawn amounts from commitments and stressed exposure. State whether “AI exposure” means a plausible disruption pathway, a material revenue dependency, or another documented classification rule.
What current evidence does—and does not—show
The Bank for International Settlements reported in its 14 July 2026 bulletin that business development companies (BDCs) had around $115 billion in loans to software firms, about a fifth of BDC lending and over 80% of their fast-growing technology portfolios. The bulletin said uncertainty about generative AI’s effects on borrower revenue had not yet affected those loans and that BDCs and their equity investors had not priced the software exposure differently at publication. It also described recently narrowed credit spreads, which reduce loss-absorbing buffers, and shared borrower pools across some large BDCs. The authors noted that low leverage and secured lending may limit spillovers. These are sector observations, not estimates of losses for any particular fund. Read BIS Bulletin 128.
#1 Best Overall
A separate Federal Reserve staff note, published 23 May 2025, examines banks’ lending to private-credit vehicles—not the underlying borrower books of those vehicles. Its sample found moderate concentration in banks’ commitments to private-credit vehicles and discussed stress if vehicles draw unused bank lines. In a hypothetical full-draw scenario, the note estimated $36 billion in increased drawdowns, 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 are modeled results for that sample and scenario, not estimates of AI-related portfolio losses. In the note’s sample of 40 publicly traded BDCs, leverage rose from about 40% in 2017 to 53% in 2024. Read the Federal Reserve staff note.
Build a look-through map of borrowers and shared dependencies
Classify each borrower using a documented rationale rather than a broad “tech” label. A software company might be exposed because its product is substitutable, because its customers are adopting AI in ways that could reduce demand, or because it depends on a particular technology provider. Track the rationale and confidence behind the tag; unknown data should remain visible as unknown, not be counted as zero exposure.
For each borrower or connected group, capture the dimensions that can connect otherwise separate loans:
Recommended Free Tools
- Sector and software sub-sector, revenue source, and the share of revenue tied to potentially AI-sensitive products or customers.
- Customer and supplier dependencies, end markets, geography, and any common technology or infrastructure dependency.
- Sponsor relationships and borrower links across funds, loan vehicles and co-investments.
- Loan amount and status (drawn, committed or otherwise), maturity, seniority, covenant package and collateral.
Aggregate borrowers and sponsors where connections create a common source of risk. A low exposure to any one borrower can still conceal a large exposure to the same end market, customer type, technology or refinancing condition.
Measure concentration from more than one angle
Report exposure shares at the level that matches the risk question: largest borrower and connected-group shares, sector and sub-sector weights, and top-N borrower shares. Use consistent exposure definitions and show the denominator. For example, a share of drawn loans is not interchangeable with a share of commitments that includes undrawn amounts.
Herfindahl–Hirschman Index (HHI) can summarize concentration across a chosen set of units, such as borrowers or sectors. On the 0-to-1 scale described in the Federal Reserve staff note, higher values indicate less diversification. State what units and exposure measure went into the calculation; an HHI for borrower names does not capture common revenue drivers. Supplement it with overlap measures for shared borrowers across funds, common sponsors, end markets and AI-sensitive revenue dependencies. No single index measures correlated risk or establishes that a portfolio is safe.
Compare credit quality and loss protection
Sector exposure alone does not show how an adverse change becomes a credit loss. Compare AI-exposed borrowers with other portfolio segments on the capacity to absorb disruption and repay debt. Keep the exposure classification separate from the credit-risk grade so that a sector label does not substitute for borrower-level underwriting.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- Cash generation: leverage, debt-service capacity, recurring versus discretionary revenue, customer concentration and liquidity runway.
- Time and refinancing: maturities, expected refinancing dependence and the availability of cash or sponsor support.
- Contractual protection: covenant headroom, collateral coverage, lien priority and recovery assumptions.
- Valuation dependence: reliance on enterprise value to support repayment or recovery, including how much downside the capital structure can absorb.
Stress linked drivers, not just individual borrowers
Run downside scenarios that connect plausible AI-related business changes to financing outcomes. Test ranges for product substitution, customer churn, pricing, growth, margins and required investment. Then combine those pressures where appropriate with higher financing costs, weaker refinancing availability, lower enterprise values, covenant breaches, weaker collateral recoveries and correlated draws on financing lines.
Show how each scenario affects defaults, recoveries, stressed losses, liquidity needs and any portfolio limits. Explain which borrowers or shared factors drive the result and identify assumptions explicitly. The reviewed sources do not provide probabilities for AI-specific scenarios; use transparent ranges rather than presenting invented precision as a forecast.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Set limits, monitoring and escalation around the portfolio
Translate the portfolio’s documented risk appetite into borrower-name, connected-group, sector, sponsor and shared-factor limits or watch thresholds. There is no evidence here for a universal percentage cap: appropriate thresholds depend on the portfolio, underwriting approach, risk tolerance and ability to monitor exposures.
Assign owners, review frequency, independent challenge and an escalation route to the investment committee or board. Monitoring triggers can include deteriorating borrower data, rapid growth in a sector or shared exposure, a limit breach, covenant pressure, spread or valuation changes, or a rising share of exposure that cannot be classified with confidence. Link contingency planning to the liquidity and financing channels relevant to the portfolio.
Interagency commercial real estate concentration guidance recommends supportable segmentation, limits and sublimits, portfolio-level management, correlation analysis, timely management information, stress testing and contingency planning. It cautions against splitting segments merely to make a concentration appear smaller. These are useful risk-management principles by analogy; that guidance concerns commercial real estate lending and does not directly regulate private-credit funds. Interagency leveraged-lending guidance likewise supports measurable underwriting standards, borrower sustainability analysis, realistic downside scenarios, and monitoring of covenants, collateral and dependence on enterprise value. Commercial real estate concentration guidance and leveraged-lending guidance.
Best Value
Keep regulatory examples within their scope
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 expressly excludes generative and agentic AI models from its scope, is non-prescriptive, and is most relevant to banking organizations; it is not an AI concentration rule for private-credit funds. Read OCC Bulletin 2026-13.
Consumer-credit disclosure requirements are another distinct 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. Read the CFPB guidance summary.
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.
