Application of Artificial Intelligence on Fraud Detection and Prevention in Deposit Taking Savings and Credit Cooperative Societies in Kenya

Abstract

Fraud remains a significant challenge for deposit-taking SACCOs, threatening financial sustainability and stakeholder trust. The increasing adoption of artificial intelligence presents an opportunity to strengthen fraud detection and prevention mechanisms. This study aimed to determine the effect of AI on fraud detection and prevention in deposit-taking SACCOs in Kenya. Anchored on the Technology Acceptance Model and Fraud Triangle Theory, it examined how AI implementation level, system complexity, employee training on systems, and regulatory compliance influenced fraud detection and prevention. It employed a descriptive research design targeting all 176 licensed deposit-taking SACCOs in Kenya, achieving a response rate of 62.5%. Primary data was collected through structured questionnaires to Chief Information Officers or Heads of IT. Descriptive statistics, correlation analysis, and multiple regression analysis were conducted using SPSS version 27. The regression model showed a strong explanatory power, with an R Square of 0.928, indicating that 92.8% of the variance in fraud detection and prevention is explained by independent variables. The results revealed that regulatory compliance had the strongest positive influence on fraud detection and prevention (β = 0.739, p = 0.000), followed by AI implementation level (β = 0.349, p = 0.000), system complexity (β = 0.308, p = 0.000), and employee training (β = 0.227, p = 0.001). These highlight the importance of aligning AI systems with regulatory standards, investing in robust implementation, and equipping employees with skills to operate AI technologies effectively. The study concludes that AI significantly enhances fraud detection and prevention in SACCOs, with regulatory compliance playing a critical role in ensuring the effectiveness of AI systems. SACCOs with higher levels of AI implementation and employee training, reported superior fraud management outcomes; it recommends that policymakers develop supportive frameworks for AI adoption, SACCOs prioritize continuous training and integration of AI technologies, and t regulatory bodies enforce compliance standards to enhance fraud detection systems. Future research could explore the long-term impact of AI on fraud detection through longitudinal studies or expand the scope to include other financial institutions. Additionally, studies on emerging AI technologies, such as machine learning and blockchain, and their ethical implications would enrich the understanding of advanced fraud prevention systems.   

Keywords:       Artificial intelligence, fraud detection and prevention, deposit-taking SACCOs, financial technology, regulatory compliance, employee competency.