Banks have made AI headway but there's a long way to go, say experts

livemint.com:

BCG. “There are use cases starting from prospecting of a customer to sales to credit underwriting to then monitoring and the collections part of it," he said. Even less mature banks have AI for underwriting, he added.

While there is a long way to go for many banks to implement AI use case across their value chain, they are at a good starting point, he said. Ramesh Lakshminarayanan, the CIO and Group Head IT at HDFC Bank, said that there are also very interesting use cases for general AI, especially in customer care. “I think handing over the prompt to the customer is a very important decision.

For example, can statements queries be handed over to customers? Why would you try and download your statements and then go to 10 different entries to figure out transactions on a particular day or location?" he said, adding that the process can be seamlessly handled by AI. Other tasks that AI can take over include banking over chat applications like WhatsApp using generative AI, or ancillary activities like writing credit access memos and legal documents, he said. Pranav Arora, the MD and Lead Applied Intelligence at Accenture in India agreed with the applications that AI can find in customer care.

“Ninety percent of us get very poor RM (relationship manager) experience (at banks). The RMs are not able to answer any second-level nuanced questions. I think banks adopting Gen AI will change that.

There is massive opportunity here," he said. Another use case for AI has been in fraud detection, according to Shiv Bhasin, chief transformation officer, IndusInd Bank. “One of the oldest use cases (of AI) has been in fraud management, especially in the transaction fraud area.