<![CDATA[The use of ChatGPT in education still faces several challenges, including misuse of technology, potential plagiarism, and academic misconduct. This study evaluates the use of ChatGPT as a task-support tool for vocational high school students using the Concerns-Based Adoption Model (CBAM) Levels of Use (LoU) framework. A qualitative case study involved 10 students from a public vocational high school in Bandung, representing diverse levels of academic achievement and skill programs. Data collection employed semi-structured interviews, direct observations, and document analysis (chat histories and assignment drafts), triangulated to support objective LoU stage mapping. Findings reveal that 50% of students are at the IVA Routine and IVB Refinement stages, 30% at Mechanical Use (III), and 20% reach Integration (V) and Renewal (VI) as AI change agents. ChatGPT provides personalized learning, flexible access, time efficiency, and instant feedback, yet raises ethical challenges, risks of overdependence, and the potential decline of critical thinking, requiring pedagogical intervention. Results indicate the use of adaptive strategies and differentiated policies based on students' LoU positions, thereby strengthening peer tutoring and ethical supervision. This ensures responsible and sustainable adoption of AI in vocational education. The novelty lies in the detailed LoU mapping for SMK students, which provides a practical typology for teachers and school administrators.]]>
