RBI guv asks banks to build human oversight, says blaming technology is 'unacceptable'

Reserve Bank Governor Sanjay Malhotra has urged Indian banks to ensure human oversight in the use of artificial intelligence (AI) systems, warning against treating algorithms or technology vendors as responsible for their actions. Speaking at the Fibac event, Malhotra emphasized that banks adopting AI must ensure that the technology supports human judgment rather than simply replacing it.

He stressed that the ultimate responsibility for a bank's decision must lie with the bank itself, not with the vendor or the algorithm. Malhotra argued that meaningful human oversight, the ability to explain, intervene, and override AI decisions, must be a design principle and not an afterthought. He urged banks to embrace AI as a capability to be responsibly harnessed, rather than viewing it as a risk to be contained.

Malhotra cautioned that increased reliance on AI could weaken human judgment and accountability. He highlighted risks such as cyber security issues, biases within AI models, and the herding effect, where a small number of financial models or technology vendors could perpetuate biases across the system. Banks are encouraged to approach data privacy as a responsibility beyond meeting legal requirements and strengthen their management of AI-related risks while working with external partners.

Additionally, banks should maintain an inventory of the AI models they use and have board-approved AI governance policies that establish clear accountability for outcomes.