Social and Pedagogical Drivers of ChatGPT Acceptance: An Extended TAM Model in Teacher Education

Generative artificial intelligence tools, particularly applications such as ChatGPT, hold significant potential to reshape the landscape of teacher education by offering opportunities for personalized feedback and adaptive learning. These tools also play important roles in teachers’ pedagogical decisions while designing their teachings in present learning environments. However, the underlying mechanisms driving pre-service teachers’ adoption of such particular tools remain under-explored. This study addresses this gap by validating an extended technology acceptance model incorporating constructivist teaching beliefs, subjective norms, facilitating conditions, ChatGPT self-efficacy, and perceived enjoyment. Data collected from 488 senior pre-service teachers in Türkiye were analyzed using structural equation modeling. Results show that ChatGPT self-efficacy and perceived enjoyment significantly predict perceived ease of use, whereas facilitating conditions do not exert a significant influence. Perceived usefulness is significantly associated with perceived ease of use, subjective norms, and pedagogical beliefs. Behavioral intention is significantly associated with perceived usefulness, subjective norms, and ChatGPT self-efficacy, with the framework explaining 54% of the variance. Given the cross-sectional nature of the data, these results offer tentative implications for teacher education. Practically, the findings highlight the potential value of scaffolded micro-teaching and institutional verification procedures for supporting responsible and equitable generative AI use in teacher education. Further recommendations are discussed in the light of teacher education procedures and policies.