Our Bureau
New Delhi
Artificial intelligence is rapidly transforming lending in India, but it comes with serious risks that could shake the country’s banking system, according to a new report from Canara Bank Research.
AI-driven lending portfolios are showing 2x higher growth and better return-on-asset potential. However, they also carry 1.5x to 2x higher credit risk, warns the Reserve Bank of India (RBI), CRISIL, and global rating benchmarks.
The speed of loan approvals has jumped dramatically. Underwriting time has dropped from 3–10 days to near-instant approvals, fueling a surge in unsecured retail loans, Buy-Now-Pay-Later (BNPL) services, and micro-credit.
Instead of relying only on CIBIL scores and financial statements, lenders now use real-time data like GST flows, UPI transactions, and digital footprints. This is helping MSMEs, gig workers, and first-time borrowers access credit for the first time.
But the risks are massive. India’s $280 billion-plus IT/BPO sector faces huge threats from AI automation. Job losses in tech hubs like Bengaluru, Hyderabad, and Pune could raise non-performing assets (NPAs) across retail and corporate loans.
The Economic Survey 2025–26 warned that AI’s impact on white-collar jobs could be worse than the 2008 financial crisis in terms of ripple effects on banking.
AI models also act as “black boxes,” making decisions without clear explanation. Bias in training data could lead to mis-pricing of risk and cause credit to freeze instantly during economic downturns.
The Canara Bank report calls for controlled AI adoption with hybrid underwriting, where human oversight is essential to cap risks.
As AI reshapes who gets credit and how risk is priced, India’s banks must balance speed and innovation with stability, before the algorithm responses go wrong.
