Private Credit AI Statistics 2026: Adoption, Use Cases and Software-Loan Exposure

1. The Short Answer: It Depends on the Definition

How many private credit managers use AI depends on what counts as use. Each 2026 survey draws that line in a different place, so each gives a different answer.

Lumonic and Armentum Partners surveyed about 150 private credit lenders and found AI tool adoption rose from about 20% to about 50% in a year (Lumonic, March 24, 2026). Oxane Partners asked 380 professionals at director level or above in the US and UK, and 40% reported live AI use (Oxane Partners, Compass 2026, June 4, 2026).

AI built into the credit process was much rarer when last measured. Lumonic's 2025 survey found "Only 5% have integrated AI into underwriting or loan management" (Lumonic, July 31, 2025).

The fair summary: about half of lenders now use AI tools, and in 2025 only 5% had AI inside underwriting or loan management. Quote the number that answers your question.

2. How to Read These Numbers

Private credit AI surveys are new, small and published by firms with their own reasons for asking. Ask five questions before you quote one.

• Who answered? Apex surveyed 105 senior leaders, mostly C-suite, and Oxane 380 professionals at director level or above. Allvue's 102 senior leaders work at private markets firms in North America and Europe, a wider group than credit managers. Senior leaders and analysts can see the same desk differently.
• What counts as use? In Oxane's survey, "87% of respondents are actively engaging with AI" while 40% report live use. Apex's "fully embedded" (85%) and Lumonic's "integrated AI into underwriting or loan management" (5%, in 2025) sound alike and sit far apart.
• Survey or portfolio data? The adoption figures record what people say. The software-loan figures in section 8 come from portfolio and loan data analyzed by the BIS, Octus and the Federal Reserve Bank of Boston.
• Primary or reported? Four sources reach this page through press coverage: PwC's survey via WealthBriefing, Armentum's comments via Creditflux, EY India's survey via Free Press Journal and PitchBook data via Forbes. Each is labeled as a report.
• When? The adoption figures were published between July 2025 and September 2026. Numbers this young move fast, so check the date before you quote one.

Each figure below carries its source and date, so it holds up when someone checks.

3. The Numbers at a Glance

Ten figures worth knowing, each with its source and date.

• About 50% of private credit lenders use AI tools, up from about 20% a year earlier, in a survey of about 150 lenders (Lumonic and Armentum Partners, March 24, 2026).
• 40% of 380 private credit professionals report live AI use (Oxane Partners, June 4, 2026).
• 85% of 105 senior leaders say AI is now fully embedded within their private credit activities (Apex Group, March 12, 2026).
• 5% of lenders had integrated AI into underwriting or loan management (Lumonic, July 31, 2025).
• Adoption of third-party underwriting technology "doubled year-on-year from around 10% to 20%" (J.P. Morgan, Machine Learning in Private Credit, September 1, 2026).
• Only 16% of credit portfolio managers view AI-enabled portfolio management, meaning monitoring, as a current priority, in PwC's survey of 120+ managers (as reported by WealthBriefing, May 27, 2026).
• Only 8% of firms rate their data maturity as high, in a survey of 102 senior leaders at private markets firms (Allvue Systems with Crisil Coalition Greenwich, February 10, 2026).
• 56% of the same respondents report ongoing dependence on spreadsheets (Allvue, February 10, 2026).
• BDCs "have lent around $115 billion to software firms," about a fifth of all their lending (BIS Bulletin No 128, July 14, 2026).
• Software made up "approximately 32% at cost and fair value" of 175 BDC portfolios at the end of the first quarter of 2026 (Octus, July 27, 2026).

The first four measure different things, which is the point of section 1. The last two come from portfolio data rather than opinion.

4. Adoption: How Many Lenders Use AI

The figures with a year-on-year comparison all show use rising fast. The single-year snapshots differ on how deep it goes.

• Among private credit lenders, "AI adoption more than doubled" in a year, from about 20% to about 50% (Lumonic and Armentum Partners, March 24, 2026).
• Creditflux, quoting Armentum on the same survey, reported use "rising from around 20% in 2025 to 50% when surveyed this year" (Creditflux, May 22, 2026).
• "Just 10% use third-party tools for underwriting or portfolio management" (Lumonic, July 31, 2025).
• Nearly all of the rise in third-party underwriting technology was "driven by AI-native tools" (J.P. Morgan, September 1, 2026).
• "87% of respondents are actively engaging with AI" (Oxane Partners, 380 professionals, surveyed mid-March to mid-May 2026).
• "less than a quarter of firms rate themselves above the industry average in AI adoption" (Allvue, 102 senior leaders at private markets firms, February 2026).

EY India's Private Credit Pulse Survey, as reported by Free Press Journal, sorts use by degree: "67 per cent of respondents use AI moderately," a further 22% are piloting, and "11 per cent said AI tools are already core to their origination and credit analysis processes" (Free Press Journal, August 22, 2026, on the June 2026 survey).

The deeper the measure, the smaller the number. AI tool use sits at about 50% in Lumonic's survey, third-party underwriting technology at around 20% in J.P. Morgan's figures, and AI at the core of origination and credit analysis at 11% in EY India's.

5. Where Managers Apply AI: Underwriting, Monitoring and the Middle Office

Underwriting draws the attention. Monitoring the loans already on the book draws much less, at least as a stated priority.

• More than half of credit portfolio managers say underwriting is where they are most likely to use AI (PwC Global Private Credit Fund Survey 2026, 120+ managers surveyed January to March 2026, as reported by WealthBriefing, May 27, 2026).
• "only 16 per cent of respondents view AI-enabled portfolio management (i.e., monitoring), as a current priority" (PwC, as reported by WealthBriefing).
• Apex's respondents see AI's value in investment decision-making (76%) and risk management (67%) (Apex Group, March 12, 2026).
• "63% are actively implementing AI across middle-office operations, yet only 4% use automated valuation models" (Apex Group eBook, April 16, 2026).
• "around 40% have been using tools for portfolio construction and optimization," Creditflux reported from the Lumonic and Armentum survey (Creditflux, May 22, 2026).

None of the figures here isolates document work, though it runs through every step: spreading financials, reading compliance certificates, drafting credit memos. The Claude for private credit guide shows what that work looks like day to day.

In these surveys, underwriting draws more interest than monitoring. That ordering matters in a year when software loans are under watch, because trouble in an existing loan shows up in monitoring first.

6. Technology Budgets

Every survey here that asked about spending found budgets heading up.

• "Four-in-five firms expect technology budgets to rise by 20% or more" (Oxane Partners, June 4, 2026).
• "Over 60% of respondents expect technology investment in operations to increase by 20% to 50%" (Apex Group, March 12, 2026).
• In the same survey, "nearly half" expect "to direct between 50% and 75% of technology budgets towards AI capabilities" (Apex Group).
• "More than three in five (62%) firms expect investment in AI to be prominent" (Allvue, February 10, 2026).

The money is coming. Section 7 shows what it runs into.

7. Data, Skills and Governance

The brakes are data, skills and trust in the output. Most of these numbers come from Allvue's survey of private markets firms, which covers more than credit.

• "only 8% of firms rate their data maturity as high" (Allvue, 102 senior leaders at private markets firms, February 10, 2026).
• "The majority of respondents (56%) report ongoing dependence on spreadsheets" (Allvue).
• 64% cite limited internal expertise among users (Allvue).
• 59% cite accuracy or reliability concerns (Allvue).
• 38% cite compliance concerns (Allvue).
• "Risk management ranks as the top operational challenge, cited by 42% of respondents market-wide and 60% of funds" (Oxane Partners, June 2026).
• Over 60% of Apex's respondents report "formal policies in place to govern the ethical use of AI in credit decision-making" (Apex Group, March 2026).
• Only 7% report cross-functional AI oversight (Apex Group eBook, April 16, 2026).

Policies are common and oversight is rare. A policy that no one owns across credit, operations and compliance stays on paper.

8. The Software-Loan Exposure Numbers

AI reaches private credit twice: as a tool lenders use, and as a risk inside the loans they hold. The second is measured from portfolio data.

• BDCs' software lending is "about a fifth of all their lending and over 80% of their fast-growing technology portfolios" (BIS Bulletin No 128, July 14, 2026).
• "Outstanding loans to SaaS firms increased from almost $8 billion in 2015 to over $500 billion, or 19% of total direct loans," by the end of 2025 (BIS Quarterly Review, March 16, 2026).
• Octus put BDC software exposure at "approximately 29% as of Sept. 30, 2025," across 155 BDC portfolios (Octus, February 26, 2026).
• By the end of the first quarter of 2026 it was "approximately 32% at cost and fair value," across 175 portfolios (Octus, July 27, 2026).
• "35% of software principal is still set to mature in 2028 and 2029" (Octus, July 27, 2026).
• "Less than 1% of software loans are behind on payments, a smaller share than for the rest of the BDC portfolios" (BIS Bulletin No 128, July 14, 2026).
• "Since October 2025, BDCs with high exposure to software firms have performed around 5 percentage points worse" than BDCs with low exposure (BIS Quarterly Review, March 16, 2026).
• "the share of BDC loans with PIK usage rose from approximately 6 percent to roughly 10 percent by early 2026" (Federal Reserve Bank of Boston, SEC filings of 168 BDCs, August 5, 2026).
• As reported by Forbes, "PitchBook data shows portfolio stress, with nearly ten percent of top BDC investments tied to distressed borrowers by mid-2026" (Forbes, October 3, 2026).

Why about a fifth in the BIS count and approximately 29% to 32% in Octus's? The two define software differently, which is why their shares differ. Quote each with its source.

So far the loans are paying. The software-loan review guide covers how to find the exposed borrowers before the numbers move.

9. Market Context: How Big Is Private Credit?

Size estimates differ because definitions differ. Give the source with any number you quote.

• The global market "reached US$4.1 trillion in assets under management (AUM) at the end of 2025" (Alternative Credit Council and Houlihan Lokey, Financing the Economy 2026, October 22, 2026).
• "Globally, private credit stands at over $2 trillion" (BIS Quarterly Review, March 16, 2026).
• PwC projects growth from $2 trillion to $3.4 trillion by 2030 (as reported by WealthBriefing, May 27, 2026).
• The Boston Fed's US-only figure: private credit "has grown to over $1 trillion in the United States" (Federal Reserve Bank of Boston, August 5, 2026).

The four figures count different things, so keep each next to its source.

10. What the Numbers Mean for a Mid-Market Lender

This section is the reading of WorkWise Solutions, which publishes this guide. The surveys and studies report; the conclusions below are WorkWise's.

1. Measure adoption in two parts, as Lumonic does: who uses AI tools, and which credit steps have AI built in. The first number flatters. The second tells you where you stand.
2. Give monitoring the attention underwriting gets. Only 16% of PwC's respondents treat AI-enabled monitoring as a current priority, while software sits at approximately 32% of BDC portfolios in Octus's count. The book you already hold deserves as much AI attention as the next deal.
3. Fix the data the AI will read. With only 8% rating their data maturity as high and 56% dependent on spreadsheets, AI pointed at a messy covenant tracker produces fast, confident mistakes.
4. Train people to check the output. Allvue's respondents cite limited internal expertise among users (64%) and accuracy or reliability concerns (59%), and analysts who learn on their own deals to check every output against its source answer both.
5. Name one owner for AI oversight across credit, operations and compliance. Over 60% of Apex's respondents have formal AI policies, and only 7% report cross-functional oversight.

For a mid-market lender, a good year looks like AI inside two or three credit steps, on clean data, with people who check it.

11. Sources and Methodology

Every figure on this page comes from one of these sources. Those reached through press coverage are marked as reported.

• Lumonic and Armentum Partners, 2026: The State of Private Credit Technology, March 24, 2026: a survey of about 150 private credit lenders.
• Lumonic, 2025: The State of Private Credit Technology, July 31, 2025.
• Creditflux, Around half of private credit originators use AI tools, May 22, 2026: reported, quoting Armentum on the Lumonic and Armentum survey.
• J.P. Morgan Markets, Machine Learning in Private Credit, September 1, 2026.
• PwC Global Private Credit Fund Survey 2026: 120+ credit portfolio managers across the US, UK, Europe, the Middle East, Asia and Australia, January to March 2026; reported by WealthBriefing, May 27, 2026.
• Apex Group, research release, March 12, 2026: 105 senior leaders, mostly C-suite, across the Americas, APAC and the Middle East; and its eBook page, April 16, 2026.
• Oxane Partners, Compass 2026, June 4, 2026: Opinium surveyed 380 private credit professionals at director level or above in the US and UK, online, mid-March to mid-May 2026. The link is the Business Wire release as republished by City A.M.
• Allvue Systems with Crisil Coalition Greenwich, 2026 GP Outlook Survey, February 10, 2026: 102 senior leaders at private markets firms in North America and Europe.
• EY India Private Credit Pulse Survey, June 2026; reported by Free Press Journal, August 22, 2026.
• Alternative Credit Council (AIMA) and Houlihan Lokey, Financing the Economy 2026, October 22, 2026: 41 managers plus 15 interviews.
• BIS Quarterly Review, Private credit's software lending meets AI disruption, March 16, 2026; and BIS Bulletin No 128, AI disruption in private credit: exposure to software firms in BDCs (Fernando Avalos, Giulio Cornelli and Egemen Eren), July 14, 2026.
• Octus, analysis of 155 BDC portfolios as of September 30, 2025 (February 26, 2026), and of 175 BDC portfolios as of March 31, 2026 (July 27, 2026).
• Federal Reserve Bank of Boston, Early Warning Signals in Private Credit? What BDC Portfolios Reveal about Emerging Risks (Fillat, Shen and Wang), August 5, 2026: SEC filings of 168 BDCs.
• Mayra Rodriguez Valladares, Forbes, October 3, 2026: reported, citing PitchBook LCD.

Each figure appears exactly as its source states it, with no sums, derived percentages or extra rounding, and was checked on October 28, 2026.

12. Where to Start

Answer the surveys' questions about your own team before you compare it with anyone else's.

1. Ask Lumonic's two questions of your own team: who uses AI tools, and which credit steps have AI built in.
2. Pick one underwriting task and one monitoring task, such as spreading a borrower's financials and checking a covenant compliance certificate.
3. Clean the spreadsheet each task depends on before AI touches it.
4. Pull your software exposure by borrower and run it through the software-loan review.
5. Name one owner for AI oversight across credit, operations and compliance.
6. Train the credit team on its own redacted files, with a written check for every output.

For the exam side, the guide to the SEC's 2027 exam priorities for private credit covers valuation, conflicts and the AI records to have ready.