What happened
On August 11, 2026, addressing the FIBAC 2026 conference in Mumbai, RBI Governor Sanjay Malhotra publicly directed Indian banks to build 'meaningful human oversight' into AI system design as a core governance principle, explicitly stating that regulators will not accept banks blaming AI models/technology for faulty decisions. He outlined a governance roadmap flagging seven AI risks (explainability, data privacy, cybersecurity, human oversight among them) and framed AI adoption as comparable in scale to liberalization/digitalization eras for the sector.
Why it matters
This is an authoritative supervisory signal from India's central bank governor establishing accountability expectations for AI-driven decisions in banking — a preview of forthcoming binding RBI supervisory guidance/rules on AI risk management for the country's banking sector, and a warning that technology cannot be used as a liability shield.
Action needed
Indian banks should document human-in-the-loop controls and accountability frameworks for AI-driven credit, fraud, and operational decisions ahead of anticipated formal RBI guidance.