GFF 2026: India could be first country to adopt agentic AI banking at scale: Winjit co-founder

India could emerge as the first country in the world to adopt agentic AI-enabled banking at scale, according to Ashwin Kandoi, Co-Founder of Winjit. Speaking to CNBC-TV18 on the sidelines of Global Fintech Fest 2026, Kandoi said India’s rapid adoption of technology puts its banking and fintech ecosystem in a strong position to lead the next phase of AI-led financial services.

Kandoi, however, said the industry is still around two to three years away from a stage where AI agents can independently manage parts of a customer's financial life.

“I think we are two or three years away from something like this. But I'm confident that, as an ecosystem in this country, we will be the first in the world towards agentic AI-enabled,” he said.

India’s technology adoption could drive AI-led banking

Kandoi said India's willingness to adopt new technologies could give the country an advantage as banks move beyond conventional digital transformation towards AI-first models.

“First in the world is what I believe, because we adopt technology better than anyone else,” he said, pointing to the widespread adoption of platforms such as WhatsApp in India as an example of how businesses and consumers have embraced technology in ways that were not necessarily anticipated by its creators.

According to Kandoi, the BFSI sector has historically been among the earliest adopters and biggest investors in new technology. He said India has already leapfrogged in fintech innovation and could continue to play a leading role as AI becomes more deeply embedded in banking.

AI adoption in banking moves beyond pilots

Winjit works with at least 25 banks across India, Africa and the US. Kandoi said that in India, the company is discussing AI with 23 of these banks, while only a smaller number have so far moved AI solutions into live production.

He said several banks are currently evaluating AI use cases and developing proof-of-concepts, with governance concerns remaining one of the key factors slowing wider deployment.

“There is some sense of governance fear around AI, but I think any new technology takes some time, considering that we are custodians of money,” Kandoi said.

He added that banks need time to gain confidence in the technology before deploying it more extensively, particularly given the sensitivity of financial data and transactions.

Customer interactions likely to see bigger AI role

While AI is already being used across banking, including in areas such as fraud detection, Kandoi expects customer-facing interactions to be one of the areas where AI adoption becomes more visible.

He said customer interactions could see a significant role for AI over the next three to six months, with banks increasingly using AI alongside human employees.

Kandoi also highlighted the importance of cybersecurity as banking becomes more intelligent and automated. He said banks and cybercriminals are effectively engaged in a continuous race, with both sides becoming more sophisticated.

“The thieves are going to get smarter, but the banks are going to get smarter, better, securing better,” he said.

Passwords and OTPs could eventually give way to adaptive authentication

On authentication, Kandoi said the industry is not yet at a stage where passwords and one-time passwords (OTPs) can completely disappear from banking. However, he said newer authentication solutions could reduce friction for customers.

Winjit is working on adaptive authentication solutions aimed at reducing dependence on traditional authentication mechanisms and improving the customer experience.

Kandoi said there is currently significant friction in banking authentication, including dependence on third-party OTP servers. He expects newer approaches to help make interactions with banking applications more seamless.

Guardrails critical for agentic AI

The move from AI that responds to queries towards AI agents capable of executing financial tasks also raises questions about accountability. If an AI agent makes an incorrect payment or another financial decision, responsibility could potentially involve the bank, technology provider, customer or the underlying AI model.

Kandoi said the industry will need to establish appropriate guardrails as agentic AI develops.

“So, we'll have to build guardrails around it, and I think the guardrails as an ecosystem, as a fintech body, as a fintech organisation, as a fintech nation is what I call it,” he said.

The transition from digital-first to AI-first banking therefore brings both opportunities and risks. While AI could make financial services more intelligent and scalable, banks will need to balance innovation with security, governance and accountability as they move towards increasingly autonomous systems.