RBI Wants Lenders to Use AI to Spot Borrower Stress Early
RBI Deputy Governor S. C. Murmu has called on lenders to make greater use of artificial intelligence and machine learning to identify early signs of borrower stress.
The message comes as NBFC lending continues to expand. Murmu cautioned that credit growth must be accompanied by strong underwriting, stress testing, early-warning systems and sound risk management.
Why This Matters
Traditionally, serious credit stress becomes obvious when repayments begin getting delayed.
By that stage, the borrower's financial position may already have deteriorated significantly.
AI-assisted monitoring can potentially help lenders identify patterns before an actual default occurs.
Depending on the information legitimately available to the lender and its monitoring systems, warning indicators may include changes in repayment behaviour, unusual account activity, rising debt exposure, deterioration in business cash flows, or other deviations from the borrower's normal financial pattern.
The objective is not simply to predict an NPA. It is to identify developing financial stress early enough for intervention to remain possible.
From Credit Appraisal to Continuous Credit Monitoring
Banking has traditionally concentrated heavily on the borrower's position when the loan is sanctioned.
Technology increasingly makes another model possible: Assess → Lend → Monitor → Detect → Intervene — rather than: Assess → Lend → Wait for repayment problems.
Early-warning systems are already used in Indian banking. AI and machine learning could make these systems considerably more sophisticated.
What Does This Mean for Borrowers?
A good credit score alone may increasingly be only one part of the lending relationship.
Lenders may place greater importance on ongoing financial behaviour, including repayment discipline, leverage, cash-flow stability and other indicators of financial stress.
For responsible borrowers, this could ultimately be positive. Better monitoring can allow lenders to distinguish between borrowers whose finances remain healthy and borrowers whose risk profile is deteriorating.
It could also create opportunities for earlier conversations about repayment difficulties, restructuring or corrective action instead of waiting until the account becomes seriously stressed.
WHY IT MATTERS
Credit monitoring is moving beyond one-time appraisal. AI and early-warning systems could allow lenders to identify financial stress earlier and intervene before repayment problems become severe.
ATLAS PERSPECTIVE
The future of lending may not be about predicting who will default. It may be about identifying financial stress early enough to prevent the default.
AI will not eliminate credit risk, nor should algorithms replace sound underwriting and human judgement. But combining financial data, early-warning signals and intelligent monitoring could fundamentally change how lenders manage borrowers after sanction.
For borrowers, the implication is equally important: creditworthiness is becoming an ongoing behaviour, not merely a score checked when you apply for a loan.
KEY TAKEAWAYS
- RBI is encouraging greater AI/ML adoption for early detection of borrower stress.
- Early-warning systems already exist across the banking ecosystem; AI could make them more predictive.
- Credit monitoring is becoming continuous rather than concentrated only at loan sanction.
- Borrower behaviour after disbursement matters increasingly.
- The best NPA may eventually be the one technology helps prevent.
SOURCE & ATTRIBUTION
Primary source: RBI Deputy Governor S. C. Murmu — 7th CII NBFC & HFC Summit, 3 September 2026. This analysis is produced by the DR Finance India editorial team. Readers should verify current regulatory guidance directly from official RBI publications.
Reserve Bank of India