By — Akshara Gupta

Abstract

In banking, many forces work silently behind the curtains that the layman is not aware of. The credit system plays a major role in the financial system of a country. It acts as a direct catalyst for promoting Economic Growth as it enables individuals to fulfill their long-term prosperity goals. Families can finance their children’s education, buy cars, or houses. Yet, extreme dependence on the traditional models of credit scoring, like formal financial history and recorded data, can lead to inconsistent results and deprive many individuals from accessing credit. That is where Automated Credit Systems play a significant role. They are driven by artificial intelligence, big data, and by algorithms that increasingly shape India’s Financial inclusion. These systems expand credit access for individuals who lack any credit history and make them worthy of getting credit. This article will explore how the AI-driven credit scoring has evolved in the Indian Banking structure and how it affects the Financial Inclusion in India. It will also critically examine how an automated credit system might lead to digital inequality.

Introduction

Financial Inclusion in India has seen an extensive escalation in the past few years. According to Global Findex, the number of Indians with bank accounts has increased to almost 85%. As Fintech gains recognition all around the world, it plays a major role in Financial Inclusion. FinTech stands for financial technology and describes technologically enabled financial innovations. From the Introduction of Core Banking Solutions to UPI in India and the rise of digital banking, all efforts have been made by the government to promote a system that combines the attributes of safety, security, and enhanced convenience. Fintech companies in India are making their base in the country with about 1300 startups and investment worth USD 5.7 billion. With the promotion of FinTech in banking and the financial sector, automated credit scoring has promoted financial inclusion, efficiency, and risk management.

Access to credit forms the cornerstone of the entire financial sector. Due to a lack of collateral, many women, people of rural households, small traders, and the informal working class remain excluded from the credit market. In response, banks, non-banking financial businesses (NBFCs), fintech companies, and digital lending platforms are increasingly using AI-driven credit scoring in their lending choices. The limitation of the traditional banking system is being handled correctly by Artificial Intelligence. By promoting real-time data points, Automated credit scoring reduces risk and extends credit to all indigenous communities.

From Instincts to Credit Scores to Automated Scoring.

In India’s early banking years, lending decisions were based on personal familiarity and references. The early 2000s brought the first national-scale system known as CIBIL, creating a standardised credit score, which any lender could refer to. The digital lending of 2010 introduced the second wave, Finn companies, and NBFC began referring to mobile payments, e-commerce behaviour, etc. AI then represents the third wave, which replaces its formulas with models that learn.

How does the automated credit scoring system work?

At its core, this system uses machine models that identify patterns and correlations with vast datasets. It uses sophisticated algorithms to analyze intricate financial history and other information to help the banks make correct lending decisions. The borrowers’ income level and employment status are also assessed. In automated credit scoring, the non-traditional parameters like rental history and social media activity are also considered. The automated credit decisioning systems create a credit score or rating that represents the borrower’s credit risk after analyzing these variables. Lenders utilize this score to determine the terms and conditions of lending, the interest rate they will pay, and whether to approve a loan. Speed, uniformity, and objectivity in the lending decision-making process are some benefits of computerized credit scoring. ML credit models are the most reliable today.

Benefits of the Automated Credit Scoring:

The three key advantages offered by the current model or the ML model include Speed, Accuracy, and Personalisation. ML models powered by API provide instant scores within milliseconds, whereas traditional models might take as long as two days, thereby promoting speed. ML models also refer to a large collection of data and then provide more accurate results. It also creates an intricate personal profile of every individual, tailoring credit offers to individuals’ needs.

How does the automated credit scoring system promote Financial Inclusion?

Due to many structural characteristics of India, the AI-driven credit system has gained significance in the country. According to the National Sample Survey Organisation, over 90% of the total workforce of India works in the informal sector, which results in limited information regarding formal income. Therefore, Indian banks are promoting AI-driven credit scoring, which promotes MSME financing as well. Banks can detect early warning signs of stress and take preventive measures using predictive credit scoring. This system is more adaptable. For example, during the COVID-19 pandemic, payment habits changed. While ML models may retrain regularly to maintain precision in both stable and volatile circumstances, traditional scorecards update slowly.

Challenges in AI-driven Credit Scoring

Regulators, lenders, and borrowers all expect the credit decision-making to be fair, apart from being accurate. There might be a challenge of exclusion and bias.

Firstly, there is unequal access to alternative data. The socioeconomic status significantly affects one’s digital footprints. While the poorest continue to be more data-poor, those who have a stable internet connection, mobile phones, and carry out digital transactions regularly will be more “visible” to algorithms.

Secondly, proxy variables that are correlated with caste, gender, location, or class are mostly used by algorithms. As an example, social inferiority may be indirectly represented in geographical data or consumption habits. Automated systems can hide discrimination behind their technical complexity instead of eliminating it.

Thirdly, there are examples of inaccurate data being referred to, which can lead to unjust exclusion. Women, migrants, and other informal workers might become prey to these inaccuracies.

Regulatory Gaps and Ethical Considerations

As each year there is dynamic evolution, the Reserve Bank of India aims to create regulations that foster innovation, but it lacks in many aspects. The initiative FinTech and Startup Acceleration (FAST) aims to connect all regulators, stakeholders, and investors, but still lacks clear guidelines related to Artificial Intelligence regulating the Credit Scores. Transparency and accountability are absent, updates are delayed, standardization is lacking, there are no efficient channels for resolving customer complaints, etc. The creditworthiness of the scores given by the Indian credit rating organizations is called into question by all these difficulties.

Indian law lacks a comprehensive framework governing algorithms, decisions related to credit scores, which is not even mentioned in the Credit Information Companies (Regulation) Act, 2005. While data protection principles are being promoted by RBI guidelines, there is no right for borrowers to obtain explanations for all the automated decisions related to their credit scores. This regulatory gap creates a vacuum, particularly as we see digital lending and banking expanding day by day.

Solution And Mitigation

Regular review and update of credit scoring algorithms can mitigate the bias. By conducting regular audits and compliance checks, the system can be made transparent. Adapting to continuous economic changes and exploring the macroeconomic indicators of the country related to the formal and informal workforce and rural and urban population can also ensure an accurate assessment of idiosyncratic risks. To prevent cybersecurity risk, multilayered measures, including encryption, excess controls, and intrusion detection systems, should be promoted.

Conclusion

The automated credit system has brought about many benefits for both the lenders and borrowers. The systems can make quicker decisions and give more accurate results. It makes the process fair and easier. As finance and technology are integrating, credit automation seems to be an important step for financial inclusion, risk management of risk and promoting accessibility to credit.

However, every technology system also brings about certain banes of exclusion, bias, and opacity, which need to be addressed, where models must provide interpretable reasons for credit approvals or even rejections. This seems to be the new future of India, wherein India should enroll the AI and algorithm accountability with the Financial regulatory governance of the country, ensuring that technology serves as an opportunity for all rather than a promoter of inequality.

About the Author:

Akshara is a third-year law Student at the Jindal Global Law School. She has an avid interest in financial law, Social Justice, and emerging technology. Passionate about researching and diving into untold stories.

Image Source – https://www.fortuneindia.com/personal-finance/build-credit-score-without-a-loan-smart-strategies-for-financial-health/124095