Can internal controls prevent fraud? Evaluating COSO with statistical and machine learning methods

Introduction: Financial statement fraud (FSF) undermines investor confidence and corporate governance systems. Weak internal controls, financial pressure, and industry norms can create conditions that facilitate fraudulent reporting. The Committee of Sponsoring Organizations of the Treadway Commission (COSO) framework provides a governance structure for evaluating internal control effectiveness, yet its ability to predict FSF remains underexplored. The study evaluates the effectiveness of COSO-based internal control indicators in detecting FSF using statistical and machine learning approaches within the theoretical framework of Institutional Anomie Theory (IAT).