Future of AI in the BFSI Industry

Future of AI in the BFSI Industry

In the early days, the financial services industry comprised primarily banks whose focus was to receive deposits and lend money to customers. It has come a long way in terms of sophisticated product offerings and the technology-centric approach to growth and innovation. It is expected that this trend will continue, with advancements in AI having the potential to continually disrupt the financial industry. Inclusive finance and expansion of global trade will receive further momentum due to technological investments in the financial services industry. Recognising this, several organisations have started making investments in AI and other related areas to remain competitive in the marketplace.

Traditional banks are threatened with the entry of fintech start-ups, which are innovating, removing the intermediaries from the system and deploying AI-powered smart systems that are overcoming the age-old practices and assumptions for servicing customers. Organisations have already deployed AI-powered tools for enhancing process efficiency through RPA and improved customer service through virtual assistants and chatbots. For instance, American Express has deployed AI-powered virtual assistant Amex Bot, which handles over 5 million customer enquiries each month and has reduced call centre operational costs by 15 per cent.48 State Bank of India’s chatbot system handles as much as 25 per cent of the queries that Google processes each day.49 According to ICICI Bank, the use of its chatbot ‘iPal’, launched in February 2017, has led to quicker responses with 90 per cent accuracy and reduced operational costs.50 Initially, several private banks had taken the lead in the deployment of chatbots for servicing customer queries.51 Lately, chatbots are being widely deployed in public-sector banks as well, and, in the coming years, they are likely to replace humans in many roles that involve handling customer interactions.

Algorithms are already fuelling trading, and the pace and the complex insights they bring to trading are simply unmatched by humans. In due course, these models will get even more sophisticated as they would also be able to bring into trading market intelligence and predictive analytics based on such information, which would help AI tools execute transactions with minimal human oversight. Goldman Sachs has created an AI-led trading platform to support trading with algorithms and improve investment strategies, which has resulted in $100 million savings annually according to their reports.52 Personalisation and hyper-personalisation would be the benchmark capability of every financial institution. With the ability that AI has to analyse vast amounts of data, it should be possible to offer customised products in banking, insurance and other allied services. Based on social media status, spends and ratings, it would be possible to assess the creditworthiness of individuals. This could form the basis for lending, which could be used in the short term as an additional parameter for determining creditworthiness and, in the long term, possibly even replace the credit scoring models. HDFC Bank has explicitly positioned Data, AI and machine learning at the centre of banking’s evolution towards Banking 3.0, framing it as the core of the bank’s future.53 The Bajaj Finserv app’s AI-powered service chat has cut service requests by 70 per cent.54 Lendingkart, the Indian fintech company providing working capital loans to small and medium-sized companies through its online platform, has been using AI-powered credit risk assessment such as financial documents, transaction history and social data, thus reducing its reliance on traditional scoring methods. As a result, as per company reports, there has been a 95 per cent automation rate in loan processing, leading to expediting loan disbursements and 40 per cent reduction in loan defaults.55 Ant Group, an affiliate of the Chinese company Alibaba, has built an alternative to the traditional credit scoring model. It uses Zhima Credit (known as Sesame Credit) to assess the creditworthiness of over 500 million users with the help of data related to mobile phone usage, social behaviour and transaction history. Its AI-powered micro-lending platform is able to process 98 per cent of loan applications with no human intervention within three minutes.56 These examples indicate the possibilities for hyper-personalisation with speed and precision resulting in customer delight. Intelligent robot advisers would be able to analyse clients’ financial scenarios, preferences and market trends to offer real-time advice to customers. Instead of depending upon analysts to recommend investment strategies, robot advisers would be able to offer such advice, saving the conversion or response time and the costs involved in hiring the services of expensive human resources. Groww, the Indian fintech firm providing investment options of stocks, exchange-traded funds and mutual funds to its customers through its online platform, has been using an AI-led robo-advisory system to offer personalised investment options based on the risk profiles and financial goals of each customer. Groww has attracted 25 million users in 2023,57 and annualised returns for its customers have improved by 8–10 per cent as compared to non-AI-managed portfolios. Similar to wealth management, investment management would also get a shot in the arm with AI tools as more dynamic information related to market status, investor goals and risk tolerances as well as sentiment analysis based on market trends and social media are factored into the algorithms. UBS has reported that there has been a 20 per cent increase in client engagement as an AI-led predictive system is enabling it to provide custom wealth management services to its clients based on their individual goals, risk appetite and market trends.58 Use of AI will get deeply entrenched in cybersecurity, fraud prevention and risk management. AI tools would be able to alert about potential breaches or frauds or discrepancies arising out of regulatory compliances on a real-time basis and help with correction of lapses or actions thereof. This would help organisations to avoid paying huge penalties on account of lack of adherence. American Express has deployed AI models on its platform that processes around 6 billion transactions per annum. These models analyse patterns and carry out predictive analytics that throws up potential frauds, which have helped reduce 50 per cent fraudulent transactions as per their reports.59 Mastercard, too, has developed its AI-led fraud detection system, which has helped reduce fraud incidents to the tune of 50 per cent for merchants using its technology.60 JPMorgan Chase has developed an AI-led risk management platform called COiN (Contract Intelligence) that helps to efficiently review complex legal documents, contracts and trade documents. What would have taken 360,000 hours of manual work to review 12,000 documents previously is now done in a few seconds.61 Although Metaverse has entered the mainstream, it is expected to pick up momentum in the coming years.62 Visa has filed patents for metaverse-related technologies. Mastercard has introduced virtual payment cards for metaverse transactions. JPMorgan Chase has launched Onyx, a blockchain-based Metaverse platform. There are still challenges around the regulatory frameworks and security apart from the critical volume required for scaling for AI-powered systems to take advantage of Metaverse. Sustainability and environment consciousness would be the driving forces with the customers, and customers would be keen on such product offerings that are centred on them. AI tools would analyse ESG data to evaluate investment opportunities. The trinity of Jan Dhan, Aadhar and UPI—the innovative initiatives of the Indian government—has transformed the lives of those who had recourse only to expensive local moneylenders until recently. Now they are able to access the banking system and a streamlined credit system in India.

As of August 2024, 45 crore beneficiaries have been covered under the Pradhan Mantri Jan Dhan Yojana,63 and as of October 2024, 50 crore UPI users are registered.64 UPI has enabled financial inclusion with 60 per cent of UPI users belonging to rural areas and 45 per cent to semi-urban areas.65 These initiatives have been path-breaking, and with the data and the access provided, it is possible to create further inclusive programmes with the help of AI models customising for various segments.

There are several examples of fintech start-ups from India that are making inclusive finance a reality with their AI-powered systems. Mobikwik has over 40 per cent customers in rural India out of its base of 100 million customers, offering them affordable micro-insurance, micro-credit and digital savings products. Using AI-powered systems, Mobikwik is able to assess creditworthiness of customers in real time, leveraging alternative data such as utility bills, mobile phone payments and e-commerce transactions.66 Another fintech company, Rupifi, has been servicing over 100,000 micro-, small and medium enterprises (MSMEs) all over India through its AI-powered system with 35 per cent of them not having prior credit history.67