1 month ago

AI Agents Face Infrastructure Hurdles in Indian Finance

AI Agents Face Infrastructure Hurdles in Indian Finance
Why AI Agents Can't Go Fully Autonomous Yet · rediff.com

AI agents are like smart helpers that can do tasks for people.

In India, banks and online shops want to use them, but the old computer systems they have were made for simple, predictable programs, not smart ones.

So, experts say we need new rules and safety checks to let AI agents work safely.

Companies are building extra software layers that act like guards, making sure the AI only does what it is allowed to do.

They also want to stop fraud and mistakes that the AI might make.

This is a big step before AI helpers can shop or pay for things on their own.

Key facts

Industry
Indian financial services and digital commerce
Main challenge
Legacy infrastructure built for deterministic software
Key requirement
New permission, authentication, audit, and liability layers
Leading companies
BigBasket, Juspay, Snapdeal, Zeta India, Wibmo
Focus areas
Fraud risk, deepfakes, hallucinations

Quotes

Sivaram Kowta

President of digital banking, Zeta India

“Banks can bring their own large language models. We provide the entire infrastructure and then you build agents on it. Agents on our stack cannot directly get access to data that it is not supposed to get.”
rediff.com
“On payments, agents acting on a consumer's behalf will need clear consent and spending controls before adoption can move beyond pilots, and that requires regulatory clarity, not just better technology.”
rediff.com

Sources

Related news