Reserve Bank of India Deputy Governor S C Murmu on Thursday called on all scheduled commercial banks and non‑bank lenders to expand the use of artificial intelligence (AI) and machine learning (ML) tools for the early detection of borrower stress. In a speech to a gathering of senior bank executives, Murmu warned that traditional credit‑risk models often lag behind rapid changes in borrowers’ financial health, leaving institutions vulnerable to a surge in non‑performing assets.
Early identification of stress signals, Murmu explained, can enable lenders to intervene with restructuring, targeted support or tighter monitoring, thereby containing losses and preserving financial stability. The RBI’s mandate to safeguard the banking system has increasingly focused on proactive risk‑management, especially after a series of high‑profile defaults that highlighted gaps in conventional underwriting. By leveraging AI‑driven analytics—such as real‑time cash‑flow monitoring, transaction pattern analysis, and predictive scoring—banks can spot deteriorating repayment capacity well before standard ratios turn sour.
ALSO READ | Apple’s First Foldable iPhone Duo Set for Tonight’s Launch with Leaked Color Lineup and Pricing
Globally, major financial institutions have already embedded AI/ML into credit‑risk pipelines, using vast datasets that include payment histories, social‑media sentiment, and macro‑economic indicators. In India, a handful of forward‑looking banks have piloted similar platforms, reporting faster flagging of at‑risk borrowers and more nuanced risk segmentation. Murmu’s remarks signal an intent to move these pilots from isolated projects to a sector‑wide expectation, aligning Indian practice with international standards.
The RBI’s push follows earlier regulatory steps that encouraged digital lending, fintech collaboration, and the adoption of advanced analytics. In 2023, the central bank issued guidelines on data‑driven credit underwriting, and in early 2025 it launched a sandbox for AI‑based risk models. Murmu’s current appeal builds on that foundation, urging banks to formalise AI/ML integration, invest in talent, and establish governance frameworks that ensure model transparency and fairness.
Nevertheless, the transition will not be without challenges. Banks must address data‑privacy concerns, ensure the quality and granularity of input data, and guard against algorithmic bias that could unfairly penalise certain borrower segments. The RBI is expected to issue detailed supervisory expectations to help institutions navigate these issues while maintaining the integrity of credit‑risk assessments.
