Exploring perceived AI substitution in future accounting frameworks: the role of psychological trust, anxiety, and cognitive adaptability.

Artificial intelligence (AI) is increasingly transforming accounting systems through automation, predictive analytics, intelligent auditing, and financial decision-support systems. While scholarly and professional debates frequently address whether AI may substitute human accountants, empirical evidence examining actual substitution remains limited. Existing studies have primarily focused on technological efficiency and adoption intentions, with insufficient attention given to the psychological factors shaping professionals' perceptions of AI's substitutive potential. This study therefore investigates how psychological trust, AI anxiety, and cognitive adaptability influence accounting professionals' perceptions regarding AI's potential to substitute selected accounting tasks within future accounting frameworks in China. A mixed-method research design integrating quantitative and qualitative approaches was employed. Primary data were collected from 512 accounting professionals across major Chinese cities using structured questionnaires. Partial Least Squares Structural Equation Modeling (PLS-SEM) was used to test the proposed relationships. Additionally, semi-structured interviews were conducted with 25 accounting professionals to contextualize and enrich the quantitative findings. The results indicate that psychological trust positively influences perceived AI task substitution, whereas AI anxiety negatively affects such perceptions. Cognitive adaptability emerged as a significant positive predictor and mediated the relationship between psychological trust and perceived AI substitution. Qualitative findings revealed that although participants perceived AI as having substantial potential to automate repetitive and analytical accounting activities, they consistently emphasized the continuing importance of human judgment, ethical reasoning, strategic interpretation, and interpersonal communication. The study contributes to AI-accounting and psychology literature by integrating psychological dimensions into understanding perceived AI substitution and suggests that future accounting frameworks are more likely to reflect human-AI complementarity rather than complete occupational replacement. These findings reflect accounting professionals' perceptions and expectations regarding AI substitution and can't be interpreted as evidence of AI's actual capability to perform accounting tasks or replace professional judgment.