This paper investigates cross-national patterns of artificial intelligence (AI) use among secondary school teachers using data from the Teaching and Learning International Survey (TALIS) 2024. The study is theoretically grounded in established models of technology adoption, particularly the Unified Theory of Acceptance and Use of Technology (UTAUT; Venkatesh et al., 2003) and the Technological Pedagogical Content Knowledge framework (TPACK; Mishra & Koehler, 2006). These perspectives emphasize that technology integration is shaped by teachers’ digital competencies, perceived usefulness, ease of use, pedagogical beliefs, and contextual support (Davis, 1989; Scherer et al., 2021). Building on research on digital media use, which has shown substantial heterogeneity among teachers (Pozas et al., 2022), this study examines how AI-related practices vary across countries. Applying multilevel latent class analysis (MLCA; Henry & Muthén, 2010), the study identifies distinct teacher-level AI usage profiles and explores how their prevalence and predictors differ internationally. Latent class approaches have proven effective for analyzing technology use patterns (Graves & Bowers, 2018), yet systematic cross-national analyses focusing on AI remain scarce. The analysis distinguishes between individual characteristics (Level 1) and system-level influences (Level 2), including digital infrastructure, leadership support, and professional development opportunities (Buabeng-Andoh, 2012; OECD, 2020). Three research questions guide the study. First, it examines whether similar AI usage profiles emerge across countries or whether some profiles are context-specific (Eickelmann & Vennemann, 2017). Second, it analyzes how the probability of belonging to specific profiles varies internationally. Third, it investigates whether profile membership is mainly explained by teacher-level factors—such as self-efficacy, attitudes toward AI, and training—or by broader structural conditions (Ifenthaler & Schweinbenz, 2016; UNESCO, 2021). System-level effects are captured through aggregated indicators from teacher and school leader reports, including AI-related policies, training accessibility, and governance frameworks. Additional national indicators such as innovation capacity and education expenditure are considered. This multilevel approach allows a comprehensive analysis of contextual determinants of AI integration. Methodologically, MLCA identifies latent subgroups based on reported AI practices such as lesson planning, assessment, adaptive learning, and support for students with special needs. Measurement invariance tests assess cross-national comparability, and model selection is guided by statistical fit indices and interpretability. Overall, the study provides internationally comparable evidence on the diversity of AI adoption among teachers and highlights the importance of targeted support strategies tailored to specific national contexts. The findings contribute to a differentiated understanding of AI integration in education and inform policy and professional development initiatives.

Buabeng-Andoh, C. (2012). Factors influencing teachers’ adoption and integration of ICT into teaching. International Journal of Education and Development using ICT, 8(1), 136–155. Davis, F. D. (1989). Perceived usefulness and perceived ease of use of information technology. MIS Quarterly, 13(3), 319–340. https://doi.org/10.2307/249008 Eickelmann, B., & Vennemann, M. (2017). Teachers’ attitudes regarding ICT in European countries. European Educational Research Journal, 16(6), 733–761. https://doi.org/10.1177/1474904117725899 Graves, S., & Bowers, C. (2018). Exploring teacher and student use of technology. Education and Information Technologies, 23(6), 2445–2464. Henry, K. L., & Muthén, B. (2010). Multilevel latent class analysis. Structural Equation Modeling, 17(2), 193–215. Ifenthaler, D., & Schweinbenz, V. (2016). Teachers’ perspectives on learning analytics. Teaching and Teacher Education, 56, 157–167. Mishra, P., & Koehler, M. J. (2006). Technological pedagogical content knowledge. Teachers College Record, 108(6), 1017–1054. OECD. (2020). TALIS 2018 results. OECD Publishing. Pozas, M., Letzel, V., & Schneider, C. (2022). Teaching and technology use: Patterns and predictors. British Journal of Educational Technology, 53(4), 842–860. Scherer, R., Siddiq, F., & Tondeur, J. (2021). The technology acceptance model: A meta-analysis. Computers & Education, 128, 13–35. UNESCO. (2021). AI and education: Guidance for policy-makers. UNESCO et al. (2003). User acceptance of information technology. MIS Quarterly, 27(3), 425–478.