ARTIFICIAL INTELLIGENCE IN FRAUD DETECTION: ARE TRADITIONAL AUDITING METHODS OUTDATED? (original) (raw)

This study investigates the effectiveness of artificial intelligence (AI) in fraud detection compared to traditional auditing methods, with an emphasis on machine learning (ML) and natural language processing (NLP) techniques. Through a qualitative meta-analysis of relevant literature and statistical comparisons, the study evaluated AI’s efficacy and limitations in auditing. Major findings indicate that AI-enhanced audits achieve a 15-30% improvement in fraud detection accuracy and reduce audit time by up to 60% compared to manual methods. These results underscore AI's potential to process vast datasets and detect complex fraud patterns beyond human capability. However, integrating AI presents challenges in data quality and regulatory compliance, necessitating hybrid auditing models combining AI and human oversight for optimal fraud detection. The study recommends enhancing data quality standards, regular AI system updates, auditor training, and the establishment of robust regulatory frameworks. Key Words: AI, Fraud Detection, Machine Learning, Traditional Auditing, Regulatory Compliance