Key points are not available for this paper at this time.
Artificial intelligence (AI) is increasingly embedded in language education, and learners' acceptance of AI, together with their multilingual learning enjoyment (MLE) and self-efficacy (SE), is considered pivotal to meaningful learning gains. This study developed an AI Acceptance (AIAC) framework tailored to multilingual language learning and examined whether MLE mediates the association between AIAC and SE. A total of 524 multilingual undergraduates from 11 Chinese universities participated. The scale was constructed and validated through Exploratory Factor Analysis (EFA) with 235 participants and Confirmatory Factor Analysis (CFA) with 289 participants to evaluate model fit and refine items. The finalized framework comprises 14 items across four dimensions - Perceived Usefulness (PU), Perceived Ease of Use (PEOU), Behavioral Intention to Use (BI), and Actual Usage (AU) - capturing learners' AIAC in multilingual tasks. Furthermore, Structural Equation Modeling (SEM) was used to examine how AIAC is associated with learners' SE, focusing on the mediating role of MLE. Results showed both direct and indirect associations between AIAC and SE, with MLE statistically operating as a partial mediator. Based on these findings, the study recommends integrating AI into supportive multilingual tasks and immersive intercultural activities, together with AI value-informed guidance and practical learning support, to help foster acceptance, promote enjoyment, and support learners' SE in multilingual contexts. The validated AIAC framework offers a concise, context-sensitive instrument for research and program evaluation in multilingual settings and provides actionable guidance for curriculum design and teacher development aimed at sustainable AI integration.
Yu et al. (Tue,) studied this question.