Key points are not available for this paper at this time.
The integration of artificial intelligence (AI) into financial auditing disrupts traditional trust-building, significantly impacting the Audit Expectation Gap's dimensions of the Reasonableness Gap (RG) and Deficient Performance (DP). Utilizing Agency and Trust Transfer theories, this study explores these dynamics through Covariance-Based Structural Equation Modeling and Multi-Group Analysis of 431 professionals in Vietnam, comprising 206 auditees (chief accountants and internal auditors) and 225 beneficiaries (financial analysts and bankers). The participants represent a high-expertise cohort, with over 87% holding post-graduate or professional certifications and more than 45% possessing over 10 years of professional experience. The findings reveal a stark polarization in trust formation: auditees rely primarily on human auditor characteristics like competence and independence, whereas beneficiaries exhibit “algorithm appreciation,” grounding their trust entirely in AI capability and objectivity. Furthermore, institutional trust acts as a double-edged sword. While it effectively mitigates perceptions of DP, it paradoxically inflates the RG by triggering an “expectation halo effect” - a finding that refutes traditional literature which typically views trust as a harmonizing mechanism. where stakeholders falsely equate AI-assisted reasonable assurance with absolute algorithmic certainty. Ultimately, trust serves as a crucial mediator absorbing these socio-technical antecedents. This study contributes to behavioral auditing literature by highlighting the paradoxical capability of trust to exacerbate unreasonable societal expectations in a digital context. The results offer urgent practical implications for audit firms, business practitioners, and standard-setters, emphasizing the need for strategic expectation management and the careful calibration of trust as algorithmic agents become central to the audit process within emerging markets.
Nguyen Thu Hoai (Sat,) studied this question.