<![CDATA[This qualitative case study examines how hospitality students perceive, experience, and strategically use artificial intelligence (AI) chatbots, specifically ChatGPT and Gemini, to practice English speaking. Prior research indicates that Indonesian hospitality students hold high awareness of English importance yet only moderate self-readiness, particularly in complaint handling and accent comprehension, revealing a gap that calls for deeper understanding from students' own perspectives. Grounded in Task-Based Language Teaching and Social Cognitive Theory, particularly self-efficacy, this study used semi-structured interviews with eight informants, four Diploma (D3) Hospitality and four Undergraduate (S1) Management students, analyzed through Braun and Clarke's thematic analysis and triangulated with document analysis of student-AI interaction transcripts. Findings reveal five interrelated themes: AI chatbots function as a low-risk practice space that lowers social speaking anxiety; readiness gains remain situational rather than generalized; a persistent gap exists between AI-simulated interaction and authentic guest-service encounters; social anxiety is partly replaced by a new epistemic anxiety concerning AI accuracy and overdependence; and students consistently position AI as a complement to, rather than a substitute for, instructor- and peer-based learning. This study contributes context-specific evidence on AI-assisted speaking practice in vocational hospitality education and offers practical implications for integrating chatbot practice into a structured, human-validated English curriculum.]]>
Copyrights © 2026
