AI agenda: However, scale alone will not suffice. PSBs will need to serve a fundamentally different customer, extend profitable banking deeper into Bharat and preserve the trust built over decades. Artificial intelligence (AI) may well be the force multiplier that enables the reinvention of public-sector banking.
The opportunity is larger than deploying chatbots or automating isolated processes. There is a need for a distinct “PSB.AI” agenda which brings together five connected shifts: Primacy, smart capacity, Bharat impact, augmented workforce, and integrity and trust. AI is one of the few capabilities capable of moving all five simultaneously.
Customer primacy: Banking is entering a new demographic era. By 2035, Gen Z is expected to account for nearly 47 per cent of our workforce and drive $2 trillion in consumer spending. This generation expects banking to be personalised, responsive, transparent and always available. AI can connect signals dispersed across transactions, service interactions, digital journeys and relationship-manager conversations to create a continuously evolving understanding of customer needs.
Productivity equation: Our analysis suggests that supporting India’s long-term growth could require a six-to-eight-fold increase in employee productivity across PSBs. Simply adding people and processes in proportion to balance-sheet growth is unlikely to be sustainable. AI can enable banks to redesign work itself. Standardised activities can be automated, analytical work can be AI-assisted, whereas judgement-intensive decisions can remain with experienced officers. In lending, agentic AI can support everything from lead generation to underwriting and disbursement, releasing between 60 and 80 per cent capacity for relationship managers and officers.
Augmenting people: With more than 90 per cent of PSB employees operating in branches, workforce transformation will be critical. The opportunity is not to substitute human judgement but to amplify it. AI can provide employees with customer context, policy guidance, knowledge support and decision assistance exactly when needed. New joiners can become productive faster; officers can carry context across geographies and relationship managers can engage customers more holistically.
India’s next bankable customer may not always resemble yesterday’s borrower. Gig workers, creators, women entrepreneurs, farmers and micro-businesses often remain outside conventional lending frameworks despite having viable economics. AI can help bridge this gap through cashflow data, UPI activity, alternative data and behavioural signals. Equally important, AI can improve the economics of serving dispersed and smaller-ticket customers through lower servicing costs and smarter collections.
Architecture of trust: For PSBs, the cost of failure is not merely a technology issue; it risks eroding trust built over decades. Trust, therefore, cannot be added after deployment. It must be designed into AI systems through accountability, governance, explainability and continuous monitoring. Institutions that scale AI successfully will be those that move quickly without compromising control.
Realising the PSB.AI agenda requires more than use cases. Banks will need a business-owned AI strategy, operating models built for scale, modern technology architecture, AI-ready data and governance frameworks that balance innovation with control. Success will depend as much on organisational choices as technology choices. AI must move from scattered experimentation to enterprise-wide value creation.
If India is to build a $30 trillion economy by 2047, it will need financial institutions capable of matching that ambition. For PSBs, AI offers an opportunity — not simply to become larger banks, but better ones: More productive, more inclusive, more customer-centric and more resilient. Those that move fastest will not just participate in India’s growth story. They will help define the future of banking itself.
The writer is national financial services leader, EY India. The column was edited for space
 "public sector bank, artificial intelligence, Banking")