Perceived AI literacy and TPACK in pre-service teacher education: mediating roles of perceived usefulness, ease of use, and value in blended learning

Background The integration of artificial intelligence (AI) into blended learning requires pre-service teachers to develop AI-related competencies and the ability to integrate technology into pedagogy. This study examines how perceived AI literacy and technological pedagogical content knowledge (TPACK) are associated with deep learning through perceived usefulness, perceived ease of use, and perceived value of AI use. Methods A cross-sectional survey was completed by 423 pre-service teachers enrolled in blended learning courses at a university in China. Data were analyzed using partial least squares structural equation modeling (PLS-SEM), with covariance-based structural equation modeling and fuzzy-set qualitative comparative analysis (fsQCA) used as complementary robustness and configurational analyses. Results Perceived usefulness, perceived ease of use, and perceived value of AI use significantly mediated the relationships between the two competencies and deep learning. All six indirect pathways were significant in both the PLS-SEM and covariance-based SEM analyses. The fsQCA revealed an equifinal and asymmetric structure: multiple, partly substitutable conditions were sufficient for high deep learning, and no single condition was necessary for either high or low deep learning. Conclusions The findings integrate the technology acceptance model and expectancy-value theory to explain how pre-service teachers' competencies and perceptions relate to deep learning in AI-enhanced blended environments. Teacher education should support technical competence alongside positive perceptions of AI's utility, usability, and value. Because the measures were self-reported and collected cross-sectionally at one university, the findings are interpreted as associational and configurational rather than causal.