State Bank of India Scales AI-Driven Collections and Early Delinquency Prediction
India's largest bank reports analytics-led advances up 29.4 percent year-on-year as its Project SARAL models prioritize collection actions at account level.
State Bank of India has deployed machine learning models under Project SARAL that predict early-stage delinquency at the account level and prioritize collection actions accordingly, according to industry coverage citing Elets BFSI published 3 November 2026. The bank reports analytics-led advances rising 29.4 percent year-on-year to Rs 96,214 crore in the first half of fiscal 2027, alongside cash-flow-based digital lending for gig workers.
The deployment is one of the few publicly quantified examples of AI impact on a major bank's credit book, combining early-warning prediction with prioritized collections operations. The approach pairs document-heavy credit workflows with machine-learned risk prioritization, a pattern directly relevant to mid-sized lenders digitizing collections and credit monitoring.
The bank credits the models with sharper early-warning signals across a very large retail and MSME loan portfolio.