State Bank of India, Bank of Baroda, HDFC Bank, Axis Bank and Poonawalla Fincorp are among lenders developing specific use cases through a mix of external experts and internal teams as they work to adopt AI to enhance business.
India’s largest lender SBI is scouting for an intellectual property rights firm to secure patents for innovations in AI and machine learning by its analytics department. Peer BoB is also seeking external experts to generate use cases for AI.
SBI wants the selected to evaluate its AI/ML (artificial intelligence/machine learning) model against the criteria for getting patents registered. The firm must check if the model introduces new features or processes not found earlier; to determine if the model provides a “non-obvious improvement or solution to a technical problem.” A bank document on 21 December seeking bids said the firm must also check the model’s practical applications and usefulness.
Also read | Generative AI, data centres to define India’s tech industries in 2025
Jayanth Kolla, partner at AI research and consultancy firm Convergence Catalyst said the need for SBI to build its own AI and ML patents is understandable, since it may not want to work with a public model from one of the Big Tech firms given the vast amounts of sensitive and proprietary data it handles. “It is also important to note that SBI’s work here does not involve Generative AI, but other aspects of classical AI development that could help in cyber security initiatives or risk mitigation,” said Kolla.
A senior executive at a Bengaluru-based AI startup said it is in the process of applying for the position. “It is not surprising that India’s largest banker is taking this step, especially seeing how Jamie Dimon has led JPMorgan Chase to become one of the leading innovators in AI too,” the executive said on the condition of anonymity.
In July, Bank of Baroda began looking to empanel partners to implement AI and Gen AI, and released additional information on it in November. “While the bank has developed multiple AI use cases and selected use cases using Generative AI technology, it is eager to extend these capabilities to other areas across its operations,” BoB said.
Also read | Bajaj Finance pivots to profitability with AI, scales back payments ambitions
Private sector banks are not behind. Top lender HDFC Bank has been working with AI for many years, enhancing decision-making and monitoring across key areas like credit risk, marketing and customer service.
“Our AI-driven customer service bots drive our IVR (Interactive Voice Response), social media and email communication and help us improve efficiency by providing self-service options to our customers. We also use ML to deliver pre-approved offers to a large section of our customer base and are also using ML models in our transaction risk monitoring platforms to mitigate frauds and prevent money mules,” said Ramesh Lakshminarayanan, chief information officer and group head of IT, HDFC Bank. On GenAI, HDFC Bank is exploring private LLMs (Large Language Models) and Copilot tools to support customer service processes, relationship manager efficiency, and software testing use cases, Lakshminarayanan added.
In August, Axis Bank tied up with the Society for Worldwide Interbank Financial Telecommunications (SWIFT) for an AI-driven pilot to check payments fraud in cross-border transactions.
Also read | Why every tech firm wants an AI-powered chatbot
Experts said the rise of AI in Indian banking was a long time coming. They said that traditionally, banks were going to tech services firms to deploy such technologies, but today, as the tech ramps up, they have become one of the biggest recruiters of engineering talent.
“Indian banks are ramping up. IDFC First Bank and Axis Bank are building their own AI IPs already, instead of buying and working on established models. This is a clear trend, and will only increase in the near future,” said Jaspreet Bindra, founder of tech consultancy AI & Beyond.
Top lender HDFC Bank has been working with AI for many years, enhancing decision-making and monitoring across key areas like credit risk, marketing and customer service.
The push for AI is not limited to top banks either. Adar Poonawalla-backed non-bank lender Poonawalla Fincorp on 23 December said it has integrated AI into human resources functions. Arvind Kapil, managing director and chief executive of Poonawalla Fincorp told Mint the lender is testing multiple targeted use cases, including cases for AI in risk management, compliance, business development and customer service.
“The plan is to have around 20-25 such areas in the next six to 12 months,” said Kapil, adding that his plan is for Poonawalla Fincorp to use AI as a collaborative tool instead of a replacement of human resources. “For us, AI is expected to enhance productivity, bring in consistency, reduce human error and give richness to the insights based on analysing large volumes of structured or unstructured data.”
And read | LLM chatbots, search engines will co-exist, says Google’s Raghavan
Many believe lenders prefer in-house AI models using their own databases due to concerns around intellectual property and copyright issues.
Kashyap Kompella, AI expert and founder of RPA2AI Research, said that building patents would undoubtedly be difficult, but there are lots of use cases, including taking cues from AI centres of excellence to improve operational efficiencies. That said, it is not quite possible to get a patent for an algorithm since they can be reverse-engineered by seeing the source code—the patents in question would thus be of applications in banking domain to reduce credit lapses, increase returns and efficiencies and more, he said.
#Lenders #inhouse #talent
#Lenders #inhouse #talent