Artificial Intelligence, Auditor Trust, and Fraud Detection in FinTech Audits: Evidence from Iraq

FinTech has shifted financial activity towards platforms that generate high-volume, high-velocity transaction data, creating new opportunities for fraud. This study examines the association between AI-enabled auditing and auditors' perceived fraud-detection effectiveness in FinTech environments, and whether this association varies with auditors' trust in AI tools. Drawing on trust-in-automation and technology acceptance research, we surveyed 214 auditors engaged with Iraqi banks and licensed electronic payment providers and tested the hypotheses using partial least squares structural equation modelling (PLS-SEM) with bootstrapped moderation analysis. The results show a positive association between AI-enabled auditing and perceived fraud-detection effectiveness. Auditor trust in AI has no statistically significant direct association with detection effectiveness, but it positively moderates the association between AI-enabled auditing and fraud detection: the estimated association is approximately 2.6 times stronger among high-trust auditors than among low-trust auditors. Trust therefore appears to act as a moderating condition rather than an independent predictor of perceived detection effectiveness. The study extends the discussion of audit technology beyond adoption to consider how trust conditions its perceived effectiveness. It also provides evidence from an emerging FinTech market in which research remains limited. For regulators, audit firms and FinTech supervisors, the findings suggest the importance of calibrated rather than uncritical trust in AI, supported by explainable tools, targeted training and clear guidance on AI-generated audit evidence.