Will the advancement of GAI diminish international students' reliance on Chinese language teachers? Evidence from SEM and FsQCA

With the rapid advancement of generative artificial intelligence (GAI), the traditional teacher-student relationship in international Chinese language education is undergoing substantial transformation. This study investigates whether and how GAI reshapes international students' reliance on Chinese language teachers (Students' Reliance on Teachers, SRT) by integrating the Push-Pull-Mooring framework (PPM) with the Information System Success Model (ISSM) and Basic Psychological Needs Theory (BPNT). Push factors are conceptualized as the technological advantages of GAI, including system quality (SQ), information quality (IQ), and service quality (ServQ); pull factors capture the satisfaction of students' psychological needs-autonomy (AUT), relatedness (REL), and competence (COM)-derived from teacher-student interaction; and mooring factors include AI literacy (AL) and subjective norms (SN). Drawing on survey data from 260 international students studying in China, this study employs a mixed-methods analytical approach combining partial least squares structural equation modeling (PLS-SEM) and fuzzy-set qualitative comparative analysis (fsQCA). The PLS-SEM results indicate that the technological affordances of GAI significantly reduce SRT, whereas psychological need satisfaction through teacher interaction remains the strongest positive predictor of SRT. AL and SN exert direct negative effects on SRT, though their moderating roles differ across push and pull mechanisms. Complementing these findings, the fsQCA analysis reveals multiple equifinal configurations leading to both strong and weak SRT, highlighting causal asymmetry and learner heterogeneity. The findings demonstrate that while GAI can substitute for teachers in information-oriented and task-based learning functions, it cannot fully replace teachers' roles in providing emotional support, motivation, and relational engagement. This study contributes to the literature by reconceptualizing SRT as a configurational outcome shaped by technological, psychological, and social factors, and offers practical implications for redefining teacher roles in AI-enhanced international language education.