Reserve Bank of India (RBI) Governor Sanjay Malhotra recently underscored the indispensable role of human oversight in the banking sector, even as financial institutions increasingly adopt artificial intelligence (AI). Speaking at the FICCI FIBAC conference in Mumbai, Malhotra stressed that while AI offers transformative potential, it cannot replace human judgment and accountability.
Understanding AI, Not Just Adopting It
Governor Malhotra articulated that successful AI adoption in banking is not merely about speed or quantity, but about profound comprehension. "The banks that will win in the AI era will not necessarily be the ones that adopt AI faster or the most. They will be the ones that adopt AI with the full understanding of what they are deploying," he stated. He highlighted that AI represents a fundamental shift in organizing and executing operations, demanding a significant change in mindset and substantial investment in technology, IT infrastructure, and reskilling.
India's Digital Advantage
India, according to Malhotra, is uniquely positioned to leverage AI due to its robust public digital infrastructure. He cited examples like Aadhaar, UPI, DigiLocker, and ONDC, noting ongoing efforts to expand and improve platforms such as the unified lending interface and account aggregator. This advanced digital ecosystem, he believes, places India at a strategic vantage point for AI integration.
Mitigating Risks: Bias, Exclusion, and Cybersecurity
While acknowledging AI's potential, Malhotra also issued a strong warning about associated risks. He specifically highlighted biases and exclusions that can be perpetuated if AI models are trained on biased data without human intervention. "If left unchecked, it will learn and perpetuate those biases. We need to be careful about these risks," he cautioned, insisting that fairness in AI must be a design requirement, not merely a compliance checkbox. Cybersecurity risks were also a significant concern, with Malhotra urging banks to go beyond mere compliance with data protection laws and prioritize robust security and data privacy measures.
Human Responsibility Remains Paramount
The Governor reiterated that ultimate responsibility must always rest with the bank, not with an algorithm or a third-party vendor. He championed "meaningful human oversight" as a core design principle for AI systems, ensuring the ability to explain, intervene, and, when necessary, override AI-driven decisions. This ensures that sophisticated models serve as tools to enhance human capabilities, rather than replacing critical human judgment and accountability.