.jpg)
RBI Governor Sanjay Malhotra has told Indian banks that blaming artificial intelligence tools for faulty decisions will not be accepted by the regulator, the bank's customers, or the institution itself.
Speaking at the FIBAC 2026 conference in Mumbai on 11 August, Malhotra called on banks to make meaningful human oversight a design principle when building AI models. He warned that as AI adoption deepens, there is a real risk that human judgement and accountability erode over time.
Malhotra was clear that responsibility for any decision must rest with the bank. He said the argument that a model made the decision could never be an acceptable answer, and that humans must retain the ability to explain, intervene and override AI systems at points where an error could cause material harm to a customer or threaten financial stability.
The Governor also asked banks to maintain a complete inventory of AI models in use, so the RBI can monitor what is actually running across the system. He said banks need board-approved AI governance policies that assign clear accountability for outcomes, not merely for the procurement of technology.
At the same time, Malhotra pushed back against the idea that AI is simply a risk to be managed. He urged banks to adopt the technology actively rather than wait on the sidelines, and said AI could do for financial judgement what UPI did for financial transactions. He flagged credit delivery, customer service, financial inclusion and fraud detection as areas where AI could transform banking.
He also identified risks that banks must actively guard against, including algorithmic bias that could disadvantage certain geographies, occupations or communities, cybersecurity vulnerabilities, and the concentration risk that arises if a small number of technology vendors come to dominate the market. Banks will also need to invest in talent, skilling and reskilling as they build out their AI capabilities, Malhotra said.
