Exploring the Determinants of Continuous Participation in Virtual International Design Workshops Mediated by AI-Driven Digital Humans

As artificial intelligence (AI) technologies and Virtual Exchange (VE) become increasingly embedded in higher education, AI-driven digital humans have begun to feature in design-oriented virtual international workshops, providing a novel context for examining learner behaviour. This study develops a structural model to examine the links between system support, interaction processes, self-efficacy, satisfaction, and international learning intention. Specifically, it investigates how perceived AI support, system ease of use, and interaction intensity influence students’ continuous participation in international learning through the mediating roles of learning self-efficacy, interaction quality, and satisfaction. Data were collected through an online questionnaire administered to undergraduate and postgraduate students who had participated in an AI-driven digital human–supported online international design workshop, yielding 611 valid responses. Reliability and validity analyses, as well as structural equation modelling, were conducted using SPSS 22 and AMOS v.22.0. The results show that perceived AI support, system ease of use, and interaction intensity each have a significant positive effect on learning self-efficacy and interaction quality. Both self-efficacy and interaction quality, in turn, significantly enhance learning satisfaction, which subsequently increases students’ intentions for sustained participation in international learning. Overall, the findings reveal a coherent causal chain: AI-driven digital human system characteristics → learning process experience → learning satisfaction → sustained participation intention. This study demonstrates that integrating AI-driven digital humans can meaningfully improve learners’ process experiences in virtual international design workshops. The results provide empirical guidance for curriculum design, pedagogical strategies, and platform optimization in AI-supported, design-oriented virtual international learning environments.