Building AI self-efficacy in EFL teachers: A mixed-methods study of professional development using an artificial intelligence-based lesson planning tool

This explanatory mixed-method sequential study examined the artificial intelligence (AI) self-efficacy in English as a foreign language (EFL) teachers using an intervention-based, two-week program centered around Diffit, an AI-driven, lesson planning and instructional differentiation tool. 35 in-service EFL teachers (80% female, M age = 38.7 years, teaching experience 2-25 years) were enrolled in the study. The AI Self-efficacy Scale (AISES) was used to gather quantitative data in a one-group pretest-posttest study, and the qualitative data were in the form of written reflections of teachers using inductive thematic coding. The results of the Wilcoxon signed-rank test showed that the outcome of AI self-efficacy between pretest and posttest was statistically different. The qualitative analysis revealed seven major themes that explained this enhancement: AI self-efficacy, efficiency and time-saving, differentiation and access, quality control and teacher judgment, barriers to implementation, student engagement, and ethics and data concerns. The research provides empirical support to AI self-efficacy as a specific variable in the study of educational technology and offers recommendations on how to approach the design of effective professional development that helps language educators to integrate AI tools in a way that is thoughtful and sustainable.