<![CDATA[This qualitative case study explored how English students at Politeknik Negeri Sriwijaya use artificial intelligence (AI) in learning English, which benefits and challenges they perceive, and how these experiences shape their attitudes toward autonomous and technology-mediated learning. Data were collected through semi-structured online interviews with 12 purposively selected students and analyzed using an interactive qualitative model of data condensation, display and conclusion drawing. Participants chose different AI tools for different language skills and turned to them mostly when tasks or concepts were difficult, while still consulting lecturers, friends, dictionaries and textbooks. They valued fast, detailed explanations, socially easier speaking practice and feedback on mistakes, but reported inaccurate or unnatural output, costly or repetitive features, limited speaking practice and worry about dependence. All participants drew the same boundary: AI is acceptable for understanding, correcting and practicing, but excessive when it replaces their own effort. No participant reported lower confidence, and views on autonomy ranged from independence-leaning to balanced and dependence-leaning. The findings extend technology acceptance and CALL perspectives by showing that the relationship between AI and learner autonomy in vocational EFL settings varies with individual disposition. Lecturers can therefore differentiate their guidance rather than rely on uniform policies.]]>
