The Unified Payments Interface (UPI) has become a central pillar of India’s digital payments ecosystem, enabling rapid adoption and supporting financial inclusion. Despite this success, UPI faces challenges related to cybersecurity, trust and equitable access. This paper analyses these issues by combining a literature review with empirical modelling of official transaction data from the National Payments Corporation of India (NPCI). Transaction value is treated as a proxy indicator of digital financial participation. Using an Ordinary Least Squares (OLS) regression model, results show that transaction volume is the most significant driver of transaction value, with an explanatory power of R2 = 0.967, while the number of banks live on UPI is not independently significant. To complement this, an Artificial Neural Network (ANN) was applied as an AI-based indicator, achieving R2 = 0.755 and confirming that non-linear effects are limited. The paper concludes with a proposed framework integrating cybersecurity, ease of use, and AI-driven solutions to strengthen adoption and inclusion.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Cybersecurity, Digitalisation and Financial Inclusion: A Case of UPI Services in India

  • Amay Doshi,
  • Mahendra Parihar,
  • Tej Bachhav

摘要

The Unified Payments Interface (UPI) has become a central pillar of India’s digital payments ecosystem, enabling rapid adoption and supporting financial inclusion. Despite this success, UPI faces challenges related to cybersecurity, trust and equitable access. This paper analyses these issues by combining a literature review with empirical modelling of official transaction data from the National Payments Corporation of India (NPCI). Transaction value is treated as a proxy indicator of digital financial participation. Using an Ordinary Least Squares (OLS) regression model, results show that transaction volume is the most significant driver of transaction value, with an explanatory power of R2 = 0.967, while the number of banks live on UPI is not independently significant. To complement this, an Artificial Neural Network (ANN) was applied as an AI-based indicator, achieving R2 = 0.755 and confirming that non-linear effects are limited. The paper concludes with a proposed framework integrating cybersecurity, ease of use, and AI-driven solutions to strengthen adoption and inclusion.