Fraud Detection in Iraqi Listed Companies: The Impact of Big Data and Artificial Intelligence

Abstract

This study aimed to investigate the effects of big data capabilities and artificial intelligence on fraud detection in companies listed on the Iraq Stock Exchange. This applied study employed a quantitative, descriptive-correlational, and cross-sectional design. The statistical population consisted of auditors and financial managers working in companies listed on the Iraq Stock Exchange, of whom 214 participants were selected through purposive and convenience sampling. Data were collected using a structured questionnaire measuring big data capabilities, artificial intelligence, and fraud detection. Reliability was assessed using Cronbach’s alpha and composite reliability, while convergent and discriminant validity were evaluated using average variance extracted and the Fornell-Larcker criterion. Data were analyzed using SPSS version 29 and partial least squares structural equation modeling in SmartPLS version 4. Structural model results indicated that big data capabilities had a positive and significant effect on fraud detection (β = 0.38, t = 6.33, p < 0.001). Artificial intelligence also had a positive and stronger significant effect on fraud detection (β = 0.49, t = 8.17, p < 0.001). The coefficient of determination for fraud detection was 0.64, with an adjusted R² of 0.63. The predictive relevance value was Q² = 0.41, indicating satisfactory predictive power. The effect size was 0.19 for big data and 0.31 for artificial intelligence. The findings demonstrate that both big data capabilities and artificial intelligence significantly enhance fraud detection in Iraqi listed companies, with artificial intelligence exerting the stronger effect. Accordingly, simultaneous investment in integrated data infrastructures and intelligent analytical technologies can improve auditing effectiveness, strengthen internal controls, and increase the timely identification of fraudulent activities.