EXPLORING TEACHER AND LEARNER PERCEPTIONS OF ARTIFICIAL INTELLIGENCE CHATBOTS FOR ENGLISH CONVERSATION PRACTICE: A MIXED-METHODS STUDY IN SAUDI ENGLISH-AS-A-FOREIGN-LANGUAGE CLASSROOMS

Resumen

With the increasing integration of artificial intelligence (AI) in language education, chatbots are gaining attention for their potential to support English conversation practice in English as a Foreign Language (EFL) settings. In this study, I investigated the perceptions of Saudi university students and EFL teachers regarding the use of AI chatbots for enhancing speaking skills. I adopted a convergent mixed-methods design involving 226 learners and 62 teachers from Saudi higher education institutions. I collected data using two adapted and structured questionnaires, which consisted of demographic questions, Likert-scale items across five dimensions (motivation, fluency, confidence, helpfulness, and risk), and three open-ended questions per group. I analyzed the quantitative data using descriptive statistics and non-parametric tests and the qualitative data thematically. Findings indicate that learners view chatbots as helpful, motivating tools that enhance their confidence and speaking fluency. Teachers, however, expressed greater concerns regarding misinformation, pedagogical control, and student overreliance on technology. I found no significant relationship between teaching experience and risk perception, highlighting that skepticism may stem from broader institutional or pedagogical concerns rather than experience alone. This study highlights the need for balanced AI integration that supports learner autonomy while addressing teacher apprehensions. It contributes to the literature on AI-enhanced language learning by providing practical recommendations for EFL educators and policy stakeholders in underexplored regional contexts.