Mumbai, Sep 10 (IANS): Agentic artificial intelligence (AI) can help banks deliver personalised financial services to thousands of customers but financial institutions will need clear frameworks around AI agents as the technology moves from recommendations to executing decisions, State Bank of India Chairman CS Setty said on Thursday.
Speaking at the Global Fintech Fest 2026, Setty said banks were entering the next phase of AI adoption, with the deployment of AI agents across financial services offering a major opportunity.
“Artificial intelligence should always augment human resources,” Setty said. He said the initial fixed costs of deploying agentic AI could be high but lower incremental costs as adoption expands could create economic efficiencies for banks.
Setty said building foundational AI models was necessary for creating strategic capabilities in artificial intelligence. Agentic AI could enable banks to provide more personalised services at scale, he added.
However, banks are introducing the technology in a market that is diverse and complex, creating challenges for financial institutions, he said.
While banks have been using machine learning (ML) models for several years, the next major opportunity lies in deploying AI agents across financial services, Setty said.
SBI has been among the early adopters of artificial intelligence and machine learning in banking and already uses conventional AI-based approaches for credit assessment and cash flow-based lending, he said.
Setty also highlighted the role of India’s digital infrastructure in enabling technology adoption. “India has successfully converted digital infrastructure into trust and then converted that trust into scale,” he said.
As agentic AI becomes more embedded in banking, institutions will need clear frameworks governing the identity and authentication of AI agents, customer consent and transaction limits, according to him.
He said accountability would become increasingly important as AI systems move beyond providing recommendations and begin executing decisions.
