Perceived AI Capability: The Determinant of Tax Evasion Intention

Abstract

This study uses personality factors as a moderating variable to investigate how perceived AI capacity and tax morale affect tax evasion intention. This study employs a quantitative survey approach using questionnaires distributed to taxpayers in Indonesia through convenience sampling and using Structural Equation Modeling with LISREL (SEM-LISREL) to examine the direct and moderating effects among the variables. The results show that both tax morale and perceived AI competence have a negative and significant impact on tax evasion intention, indicating that lower intents to evade taxes are linked to better tax morale and stronger perceptions of AI capabilities. The negative correlation between perceived AI competence and tax evasion intention is somewhat mitigated by openness to experience, one of the five personality qualities. The negative connection between tax morale and tax evasion intention is considerably strengthened by agreeableness and receptivity to new experiences, suggesting that personality traits affect how well moral and technological systems reduce tax evasion intention. The findings extend the implementation of the Theory of Planned Behavior by demonstrating that technological perceptions, tax morale, and personality characteristics jointly contribute to explaining tax evasion intention. Practically, tax authorities should strengthen AI based tax monitoring and detection capabilities while simultaneously fostering taxpayers' tax morale and considering individual personality differences in designing tax evasion strategies. By combining perceived AI capabilities, tax morale, and the Big Five personality traits into a single framework to explain tax evasion intention in the setting of Indonesia's digitalized tax administration, this study adds to the body of knowledge.