Determinants of AI Distrust Among Students in Innovation Learning

Abstract

This study investigates the effects of distrust in data protection, functional distrust, and distrust of AI replacing humans on students’ distrust of Artificial Intelligence (AI) in an innovation learning context. This study employs a survey-based approach and partial least squares Structural Equation Modeling (PLS-SEM) with bootstrapping to examine the proposed hypothesis. A total of 104 responses from innovation students at the University of Surabaya were collected using a structured questionnaire. Findings reveal that all three hypotheses were supported. Functional distrust about performance skepticism emerged as the strongest predictor of AI distrust, followed by data privacy concern and human superiority belief, with the three variables collectively explaining a substantial proportion of the variance in AI distrust. Findings suggest that educators and AI developers should address students’ capability skepticism and privacy concern to foster greater AI acceptance for innovation in higher education.