Artificial Intelligence and Fraud Detection Efficiency in Commercial Banks in Nigeria: Evidence from Calabar Metropolis, Cross River State
Abstract
This study investigated the effect of Artificial Intelligence (AI) on the efficiency of fraud detection in commercial banks in Nigeria, focusing on the Calabar Metropolis, Cross River State. A descriptive survey design with a quantitative approach was adopted, and primary data were collected from 124 valid banking professionals across Fraud Forensics, IT/Cybersecurity, Risk Management, and Digital Banking Operations departments using the AI and Fraud Detection Survey (AIFDS) questionnaire. Four AI dimensions, machine learning algorithms, deep learning systems, AI-driven operational processes, and IT infrastructure quality, served as independent variables, with fraud detection efficiency as the dependent variable. Data were analyzed using descriptive statistics, Pearson correlation, simple linear regression, and multiple regression via SPSS Version 29. Findings revealed that all four AI dimensions have significant positive effects on fraud detection efficiency: machine learning algorithms (β = 0.647, R² = 0.419, p < 0.01) emerged as the strongest individual predictor, followed by AI-driven operational processes (β = 0.638, R² = 0.407, p < 0.01), deep learning systems (β = 0.619, R² = 0.383, p < 0.01), and IT infrastructure quality (β = 0.573, R² = 0.328, p < 0.01). All four null hypotheses were rejected at the 0.01 significance level. Jointly, the four dimensions explained 58.4% of the variance in fraud detection efficiency (Adjusted R² = 0.571, F(4,119) = 41.77, p < 0.01), with the regression equation FDE = 0.384 + 0.271(MLA) + 0.238(DLS) + 0.254(AID) + 0.196(ITF). The study concludes that AI is a statistically significant and strategically indispensable determinant of fraud- detection efficiency in Nigerian commercial banking, and recommends that banks adopt an integrated, multi-dimensional AI strategy, prioritize IT infrastructure investment, and institutionalize real-time, AI-powered transaction monitoring to curtail escalating financial fraud.