Indian lenders are moving AI beyond pilots, but bankers warn that weak data, cyber threats and opaque decision-making could undermine large-scale deployment.
Indian banks are moving from small-scale artificial intelligence trials towards wider deployment, but executives at the Business Standard BFSI Insight Summit argued that the next stage will only work if lenders pair ambition with discipline.
R Subramaniakumar, managing director and chief executive of RBL Bank, said the Reserve Bank of India has made its position clear that banks must move beyond experimentation and into implementation. He said Indian lenders have already identified promising uses for AI, but success will depend on the quality of the underlying data, as well as the volume and consistency of that information. Rajiv Anand, managing director and chief executive of IndusInd Bank, struck a similar note, saying banks need to stop thinking in terms of pilots and start focusing on large-scale execution around specific business problems.
The comments come as Indian banks appear to be further along in AI adoption than the debate over scale might suggest. A 2026 Zeta survey reported by multiple industry outlets found that seven in 10 Indian banks have already put AI into production, while about 30% have achieved scaled deployment. Even so, the same survey pointed to a persistent gap between proving that AI works in one area and reproducing that success across an entire institution, with security, control and infrastructure weaknesses still slowing roll-outs.
That divide was echoed by bankers at the summit, who said AI should not simply be layered on top of existing workflows. Anand said the technology should be used to reimagine customer journeys, not merely automate old ones. He pointed to examples such as loan applications that could be processed through messaging platforms like WhatsApp, where customers upload documents, the system reads the material and a decision is triggered more quickly than in a branch-based process. Banks are also testing AI agents and voice-based tools to handle routine tasks and lift productivity.
The potential benefits, however, bring a heavier burden of oversight. The bankers said ownership of decisions must remain with the bank, even where models are built externally or adapted from specialist systems. They also stressed that AI-driven decisions must be transparent and explainable to both regulators and customers. That requirement is becoming more important as AI is used not only for customer service and fraud monitoring, but also for credit assessment, complaint handling and other functions that can affect financial outcomes for borrowers.
Cybersecurity and governance are emerging as central issues in that shift. A recent Boston Consulting Group report on India’s financial sector said AI is changing the cyber threat landscape by allowing attacks to evolve faster and exploit systems at machine speed, with 76% of Indian BFSI chief information security officers ranking AI-enabled attacks among their top priorities for 2026. At the summit, bankers said similar risks mean lenders need stronger data-layer protection, traceability, auditability and resilience if models fail or become obsolete. V Vaidyanathan, managing director and chief executive of IDFC First Bank, said data practices are as important as the technology stack itself, and estimated that most banks are already spending about 10% of operating expenditure on technology.
The wider policy backdrop is also changing. Reserve Bank of India governor Sanjay Malhotra has argued that AI should be used to widen credit access, including by helping banks assess first-time borrowers and small businesses that may not fit conventional scoring models. That creates a dual challenge for lenders: to use AI aggressively enough to improve lending and customer service, while keeping clear controls in place so that automation does not weaken accountability or create fresh risks. According to the bankers at the summit, the winners will be those that combine scale, clear use cases and robust governance rather than chasing AI for its own sake.
Disclaimer: This article is intended to inform and educate, not to recommend or endorse any financial product, investment or strategy. Please consider your own financial circumstances and seek professional advice where appropriate before making financial decisions.
