6 min readNew DelhiOct 30, 2024 07:19 AM IST
First published on: Oct 30, 2024 at 12:04 AM IST
‘iPal’, an omnichannel bot fielded by ICICI Bank – the country’s second largest private bank by assets – tackles general banking queries from customers, including transactions such as bill payments and fund transfers, helping users navigate within the bank’s app or website or extending personalised product offers. ‘ILA’, SBI Cards’ interactive live assistant, too can aid customers through the process of filling out a new application for a credit card, and assist in customers gaining access to reward points or deactivating international usage.
Dovetailing artificial intelligence (AI) and machine learning (ML) tools progressively into core operations, India’s bank sector players are leveraging AI to improve customer experience, pare costs, manage risks and drive growth through chatbots such as ‘iPal’ and ‘ILA’. In some cases, these tools are also being used for screening of potential customers for access to loans, something that lends itself to greater regulatory scrutiny, given how foolproof these evaluations are and the potential of inherent biases. But as lenders increasingly deploy these tools, the usage of AI related keywords in the annual reports of private sector banks has evidently shot up nearly six-fold in the 2022-23 reports as compared to 2015-16 levels, while even for public sector banks, the emphasis on technologies such as AI in their annual reports has increased more than three times between 2015-16 and 2022-23.
Does this rapid embrace of AI tools pose a risk? The Reserve Bank of India does seem to think so, with India’s central bank governor asserting that “the heavy reliance of AI can lead to concentration risks, especially when a small number of technology providers dominate the market”. Shaktikanta Das said at an industry event here on October 14 that the use of these tools could amplify systemic risks as failures or disruptions in these systems may cascade across the entire financial sector.
While he did not elaborate on the specific trigger for the comments, Das did allude to the growing use of AI introducing new vulnerabilities, such as increased susceptibility to cyberattacks and data breaches and AI’s opacity making it difficult to audit or interpret the algorithms that drive decisions, potentially leading to “unpredictable consequences” in the markets. “In the ultimate analysis, banks have to ride on the advantages of AI and Bigtech and not allow the latter to ride on them,” he said.
Currently, most Indian banks are using AI to primarily improve customer experience, and drive growth through chatbots and customised banking. In some cases, they are being put to use for screening of new customers for access to products such as loans. What is clear is that the use of AI in India’s banking sector is primarily being led by the private sector lenders, and among them, the larger institutions.
An analysis of annual reports for 32 commercial banks for eight years, starting from FY16 to FY23, has showed that initially (FY16), public sector banks were proactively considering AI/ML and related technologies, showing a similar mention of AI related keywords (AI score) compared to their private sector counterparts. However, during the period from 2016 to 2021, the AI-related word count in the annual reports of private sector banks has picked up, the analysis by a team from the RBI’s Department of Economic and Policy Research showed. This, according to the study, may be due to a combination of recognition of additional use cases of existing AI based technologies along with more agility in adopting the newer and advanced AI techniques and models.
The usage of AI related keywords in the annual reports of private sector banks increased around six-fold in 2022-23 reports as compared to 2015-16 level. Even in the case of public sector banks, the emphasis on new age technologies like AI in their annual reports has increased more than three times between 2015-16 and 2022-23. The word cloud distilled from the annual reports reveals interesting insights with most banks focussing on automation which may be due to a push for efficiency gains and reduce human interventions, with data analytics another major thrust area with possible usage in fraud detection and predictive analytics.
Across all models, the RBI study has found that the AI score is positively related to the asset size of the banks as evident from positive and statistically significant coefficients, suggesting higher adoption by larger banks. Also, larger banks are likely to achieve higher net gains from adoption of such technologies and data integration, thereby increasing the motivation for adoption of AI, it said. It may also be indicating that the adoption of technologies such as AI is relatively difficult for smaller banks due to larger fixed costs and absence of economies of scale.
Analytics use in banks
Customer facing bots: for fielding general banking queries from customers, helping users navigate within the bank’s app or website or extending personalised product offers
Pre-approved loans: Existing customers deemed eligible are informed through SMS/email and instant loans are sanctioned through internet/mobile banking without need for documentation and even without visiting the branch.
Eligibility tests: Transactions in existing accounts are scrutinised and qualifying customers identified. Preapproved offers are then sent through internet banking. If the customer accepts the offer, loan will be approved after minimal documentation.
Early warning System: Indications of stress in standard accounts are identified early and alerts then sent to operating staff. Early identification helps bank to take corrective action and curb slippages, thus mitigating credit risk.
Customer facing bots: for fielding general banking queries from customers, helping users navigate within the bank’s app or website or extending personalised product offers 
Share
