Artificial intelligence and machine learning are transforming private credit by automating the grunt work of analysis.
Where analysts used to spend hours manually pulling data from messy, non-standardized documents, machine learning can now extract and standardize that information in seconds — one JPMorgan (NYSE:JPM) report cites cutting per-company data extraction from 45 minutes to 30 seconds.
Predictive models, the New York-based bank claims, could help credit analysts spot potential defaults before they happen.
"The probability of default is exactly what we’re trying to solve," Sarah Gang, global head of underwriting, credit financing at JPMorgan, wrote.
For a portfolio of 100 companies, AI could potentially generate a probability-of-default assessment for each borrower. Analysts could then begin with the companies showing the highest risk and investigate further.
Machine Learning Augments Analysts
The push comes as the private credit industry balloons. Preqin estimated that the sector’s assets under managment would reach roughly $2.28 trillion in 2025 (JPMorgan had it at about $3.5 trillion in 2024).
Either way, portfolios continue to grow larger and more complex as lenders seek new ways to process information. Machine learning, experts argue, allows analysts more time to focus on the actual credit decision rather than manually assembling the information needed to make it.
AI could also expand the types of information lenders use to monitor borrowers. Alternative data, including payment flows, supply-chain activity, receivables performance and broader industry signals, can provide clues about a company’s liquidity, operating health and refinancing risk.
The goal is not simply to approve loans faster. Instead, JPMorgan said these datasets can help lenders identify signs of deterioration earlier, particularly in private credit portfolios that are less liquid and more difficult to price.
Adoption of AI-enabled underwriting technology is already accelerating. Kevin Hsu, chief executive of Lumonic, a PitchBook company, said a survey of roughly 150 private credit lenders found adoption of third-party underwriting technology doubled year-over-year, from about 10% to 20%, with nearly all of the increase driven by AI-native tools.
Yes, But…
At least one expert expressed caution and emphasized the importance of human oversight — particularly when artificial intelligence makes predictions that could influence lending decisions.
Gabriele Butti, Global Head of Credit Quantitative Research at J.P. Morgan, pointed out that as new groups of private credit analysts increasingly rely on AI capabilities, there may be a dearth of future experts.
“My concern is not even a short-term one,” Butti said. “Let’s say these models can be wrong right now. But fast forward any number of years—who’s going to have the expertise to recognize a wrong output? When today we speak about the human in the loop, we implicitly assume that the human has the expertise to spot wrong outputs. But the human has that expertise because they created that in a world before the existence of this model. Who’s going to have the expertise 15 years from now?”
Photo: Skorzewiak / Shutterstock
Login to comment