Overview
- RBI Governor Sanjay Malhotra said Tuesday that banks must retain meaningful human oversight and that ultimate responsibility for decisions rests with the bank, not the algorithm or a vendor.
- The central bank told lenders to prepare board‑approved AI governance, keep a full inventory of models in use, run red‑teaming and stress tests, and ensure explainability and human override where errors could harm customers.
- The RBI will use a proportional, consultative approach that supports sandboxes and shared utilities such as MuleHunter and a proposed Digital Payments Intelligence Platform to test and strengthen AI tools.
- Bank leaders including SBI chairman Challa Setty are pushing AI beyond urban retail to expand rural credit and small‑business lending, while the RBI said machine‑learning is essential to detect fast, API‑driven fraud.
- Regulators warned of concrete risks from opaque models, data bias, vendor concentration, third‑party dependencies, data privacy and cyber attacks, and said India’s public digital infrastructure gives the sector a strategic advantage if banks invest in tech and skills.