The Indian banking system has been grappling with large value loan frauds for some time now. Over the last four years, loan frauds have amounted to Rs 77,521 crore across public and private sector banks, foreign banks, small finance banks, local banks and other financial institutions. In 2018-19, loan frauds accounted for 90 per cent of the total while in 2019-20, they accounted for 98 per cent of total scams.

Recently, RBI constituted a working group on digital lending to study regulated lending activities and unregulated players. This move was in response to fraudulent digital loan apps charging very high interest rates and using extortion to get repayment. Some of these unauthorised apps were infringing on borrowers’ privacy. They collected phone contacts from the borrower’s device, read their texts, and requested unnecessary microphone access to record audio.

Needs of the hour are new-age technologies that protect user data and minimise human intervention. These solutions should also improve current inefficient methods of collecting large amounts of data and mitigate loan fraud risks. Credit rating driven by technology can help financial institutions mitigate risks early in the process. Technologies like Artificial Intelligence (AI), Machine Learning (ML), blockchain and big data analytics can allow a more accurate and extensive credit risk profiling.

This is where partnerships of fintech with banks and other financial institutions become crucial. Fintechs can leverage and integrate technology into the lending framework to make it transparent and fraud-proof.

Leveraging actionable data insights

Banks collect large amounts of unstructured consumer data that barely holds any value in the decision-making process. Artificial Intelligence (AI) can mould this data into actionable insights.

AI-powered tools can study consumer data and find useful patterns that can drive real-time decision-making. These tools utilise data collected from traditional sources and leverage alternate data like borrowers’ social media activity, daily transactions, utility payments, employment history, education, and more. It is imperative to ensure that data collection is consent-based.

The more data the tools have to analyse, the better the insights are. So, they can create loan distribution models that reduce risk and build trust. Alternate data can also help provide customised credit solutions based on the borrower’s situation and needs.

AI-based learning algorithms keep improving their ability to label and segregate data over time and as they are provided with more data features. With time, they can offer accurate borrowers’ profiles and a better picture of their creditworthiness, thus alleviating fraud concerns.

Minimising human intervention and bias

Integration of AI means the removal of human intervention. Decisions made by humans are influenced by their biases which can result in loan frauds. AI-driven tools run the available data against a set of rules to determine the borrower’s acceptability. This also allows for faster credit risk assessment which leads to lower risk for the financial institutions.

As data more and more data is onboarded online, hackers are finding new ways to gain access to it. Lenders might find it always challenging to be entirely sure of a borrower’s identity when communicating virtually.

Here, the use of biometrics, along with two-step authentication, can prove to be useful. Voice identification or video-based identification will ensure that the borrower’s data is safe and protected and will eliminate human error chances.

Blockchain, due to its transparent and immutable nature, can also disrupt the lending ecosystem. It can track frauds, payments and disbursals in real-time. Blockchain also eliminates the need for lenders and borrowers to meet in-person for thorough and time-consuming identity verification. Decentralisation integrated with verification processes can create digital IDs that cannot be tampered with and can be used to validate documents. Due to blockchain’s immutability, data cannot be misrepresented, which reduces the risk of fraudulent transactions.

COVID and post-COVID scenario

The pandemic has wreaked havoc, especially on small enterprises. Small businesses are and will require personalised loans to start recovering from the grave economic impact of COVID-19. Additionally, due to social distancing being imperative, the demand for loan disbursal will primarily be through digital, no-contact means, increasing credit risks.

Digital-first approach to lending is the call of the hour. Adoption of technologies like AI needs to be more than an afterthought. These technologies will lower the time required to determine a borrower’s creditworthiness, minimise paperwork, and allow remote, secure and seamless disbursal of credit. Due to the use of alternate data, lenders will understand a borrower’s ability and intent to repay and reduce any potential credit risks.

Technology to stay secure and relevant

Many organisations have already adopted AI and blockchain technologies in their risk-detection processes. They’ve increased their profitability as AI algorithms lower underwriting and fraud detection costs due to their intelligent decision-making. Besides being beneficial to financial institutions, digital experiences that involve AI, ML, blockchain, and other technologies provide quick results that are also customer-centric. In an increasingly digital world, banks and other financial institutions must prioritise technology to stay relevant.

The author is the Co-founder & CEO of Decimal Technologies