Speaker
Description
Frequent personal use of AI does not necessarily prepare pre-service EFL teachers to guide its use in classrooms. This exploratory study addresses this issue through a descriptive survey of 26 English education students at a Vietnamese university. The survey mapped participants' AI use across five domains: personal language learning, lesson planning, materials design, assessment, and learner guidance. The findings indicate a clear contrast between frequent personal AI use and limited pedagogical agency. All participants reported no formal AI coursework, and 22 reported using ChatGPT daily. AI was commonly used for idea generation, lesson planning, materials design, and personal language support. However, only three participants reported using AI to help learners work with AI themselves. Participants regarded prompt design as a pedagogical decision and agreed that AI-mediated teaching requires new professional knowledge. Their confidence was weaker, however, when designing learner-evaluation tasks or setting boundaries for AI use in assessment. These findings suggest that personal AI fluency and pedagogical AI agency are distinct competencies requiring distinct preparation. On this basis, the study proposes a four-component AI literacy framework for pre-service EFL teachers: technical use, critical evaluation, learner guidance, and ethical boundary-setting. The framework positions AI literacy not as tool competence alone, but as a form of professional agency. For teacher education, this means treating AI literacy as part of lesson design, learner support, and assessment decision-making, rather than as an optional technology skill.
