What happened
On August 19, 2026, Reserve Bank of India Governor Sanjay Malhotra, speaking at FIBAC 2026 ('A Vision for Responsible AI, Resilient Banking'), directed banks to maintain a complete inventory of AI models in production, establish board-approved AI governance policies with clear accountability, ensure explainability for decisions materially affecting customers, conduct red-teaming/stress-testing of AI systems, and maintain meaningful human oversight where errors could cause material harm. He stated banks retain ultimate responsibility for AI-driven decisions and signaled RBI plans for a forthcoming approved AI governance policy and testing sandbox.
Why it matters
This is authoritative supervisory guidance from India's central bank/prudential regulator directly shaping AI governance expectations for all regulated banks in one of the world's largest banking markets, ahead of formal RBI rulemaking on AI governance and a planned regulatory sandbox.
Action needed
Banks operating in India should begin building AI model inventories, board-level AI governance policies, and red-teaming/explainability processes ahead of RBI's anticipated formal AI governance framework and sandbox.