The use of ChatGPT and other generative AI tools in education has increased the need to assess how these tools affect students academic performance, particularly in areas such as problem-solving and ethical awareness. However, there is a lack of a comprehensive instrument to assess these areas among Malaysian Technical and Vocational Education and Training (TVET) students. This study addresses the gap by developing and validating a questionnaire to evaluate the responsible use of ChatGPT among TVET students. The questionnaire was constructed in three phases: (i) questionnaire development, (ii) content validation by four experts using Aikens V (threshold ≥ 0.83) and (iii) pilot testing with 43 purposively sampled TVET students to assess internal consistency. Expert evaluations of item relevance, clarity, and simplicity (RCS) led to minor items refinement. The pilot results showed high reliability (Cronbachs α > 0.80) across all sections. Due to the limited sample size (n = 43), exploratory or confirmatory factor analysis was not conducted. Instead, item-level performance was evaluated using descriptive statistics, corrected item–total correlations, and Cronbachs alpha, which provided sufficient evidence for validation. The validated instrument serves as a foundation that offers a reliable tool for full-scale studies and lays a strong foundation for future research on AI-assisted problem-solving and ethical awareness in Malaysian TVET programs.