EXPLORING THE ROLE OF AI-DRIVEN CREDIT SCORING SYSTEMS ON FINANCIAL INCLUSION IN EMERGING ECONOMIES

  ABSTRACT: Currently, AI adoption is speeding up across a broad range of industries, and finance has been among the most significantly affected industries by implementing the AI concept. One of the revolutionary cases that has emerged is the creation of new sophisticated artificial intelligence based credit scoring models, which use innovative machine learning and massive amount of data on individuals’ or companies’ credit risks. These technologies are particularly useful for facilitating growth in emerging markets due to institutional problems that include weak structures in the provision of financial systems, conventional credit scoring processes that lock out significant populations, and poverty is rife in the developing world. Conventional credit-rating approaches only take into account financial history, employment records, bank statements and credit history which many people in emergent markets do not possess. This puts cost access barrier to informal employment, small business and rural dwellers consequently perpetuating poverty and exclusion. AI-based credit scoring, for example, has the potential of not only using the traditional sources of credit data, such as the borrower’s identity, payment history and credit card usage, but also mobile payment history, utility bill payments, social media activity, and geographical information. With the help of these non-trivial data sources, AI systems open more detailed and multifaceted credit histories, thus providing access to finance to such population groups to which it was impossible to assign a score before. Algorithms based credit scoring have revealed powerful application but their application is not without some drawbacks. There are significant questions about the explicit ethical problem of how AI can be supported for decision making for financial goals. This usually happens when an algorithm is programmed in such at manner, or built from data in which, various injustices are already embedded. Analyzing the theoretical and practical aspects of AI credit-scoring systems is the focus of this paper.