Reserve Bank Governor Sanjay Malhotra on Tuesday urged banks to maintain “meaningful human oversight” while designing and deploying artificial intelligence (AI) models, stating that responsibility for decisions must ultimately rest with banks and not algorithms or technology vendors. Addressing FIBAC 2026, the annual banking conference organised by FICCI and the Indian Banks’ Association, Malhotra said AI should augment rather than merely replace human judgment and urged lenders to adopt the technology responsibly. He warned that excessive reliance on AI could erode human judgment and accountability, and that banks must retain the ability to explain, intervene and override AI-driven decisions, particularly where errors could materially harm customers or affect financial stability. Malhotra flagged cybersecurity threats, model biases and the risk of “herding” if banks rely on a limited number of AI models or technology vendors, as such concentration could allow biases to spread across the financial system. He asked banks to treat data privacy as more than a legal compliance requirement and strengthen risk management when working with external technology partners. AI systems should be capable of explaining the rationale behind their recommendations, he said. He called on banks to maintain a complete inventory of AI models in use and establish board-approved AI governance policies with clear accountability for outcomes, adding that meaningful human oversight should remain a core design principle wherever AI-related errors could cause significant harm to customers or pose risks to financial stability.