Agentic AI is emerging as one of the key technology themes for financial services as banks look beyond generative AI tools that answer questions or assist employees towards systems that can carry out parts of a workflow on their own.
Global Fintech Fest (GFF) 2026 has identified agentic AI as one of the technologies that could reshape finance, with potential applications across fraud detection, compliance, underwriting, customer service and personal finance management.
Traditional automation typically follows predefined rules. Agentic AI goes further by allowing AI systems to assess a situation, determine the next step and carry out a sequence of actions towards a defined outcome.
From assistance to execution
Take a loan application. It can involve customer onboarding, document checks, credit assessment, underwriting and compliance before a decision is reached.
Live Events
An agentic system could potentially coordinate parts of that process by collecting information, checking it against defined policies, routing documents and escalating exceptions to an employee.
Similarly, in fraud monitoring, an AI agent could identify an unusual transaction, analyse related signals and initiate the next permitted action within limits set by the financial institution.State Bank of India has been exploring agentic workflows across areas including risk management, underwriting, personal finance management, customer onboarding and internal reporting.
“Any process you pick up, I think there is room for an agentic workflow to at least supplement it, if not take over it completely, within a set of guardrails given the risks,” SBI deputy managing director and head of digital banking and transformation Nitin Chugh had told ET.
Chugh said that in the future, core banking can be reimagined with agentic AI, adding that adoption would accelerate once the RBI comes out with clearer guidelines.
HDFC Bank has also been looking at AI agents for customer service, back-office operations and personalised services.
Why finance is a natural fit
Banks and financial institutions process large volumes of transactions, documents, customer interactions and regulatory information every day.
Many of these processes involve repetitive steps and clearly defined policies, making them suitable for systems that can coordinate several tasks rather than automate only one.
Early adoption, however, is likely to focus on lower-risk internal processes.
Rohit Pandharkar, partner at EY India, had told ET that initial use cases could include internal document verification and workflows where AI agents coordinate tasks currently passed manually between departments.
“It is likely that agentic AI will be initially implemented more on internal facing use cases like the steps behind issuing an extra cheque book or verification of documentations internally,” Pandharkar had said.
The RBI in December last year set up a committee to develop a Framework for Responsible and Ethical AI (FREE-AI) adoption in the financial sector.
Follow GFF 2026 with ET for live updates, insights and key developments.
Add 
as a Reliable and Trusted News Source