ChatGPT adoption and responsible use in Vietnamese transnational education students’ academic writing

Abstract

This study examines how cognitive and affective factors drive Vietnamese transnational education (TNE) students' intention to use ChatGPT for English academic writing, and investigates how information verification and responsible use enhance English writing proficiency (EWP). Survey data from 501 TNE students across three major Vietnamese cities were analysed using confirmatory factor analysis and structural equation modelling. The conceptual model integrates the technology acceptance model, the pupils’ attitudes toward technology short questionnaire (PATT-SQ) model and Mitcham’s philosophical framework. Insights from an expert survey of six lecturers supported the interpretation and practical implications. Cognitive factors (perceived usefulness, perceived ease of use and perceived risk) significantly influence ChatGPT usage intention, whereas affective factors do not. Perceived risk negatively impacts intention, reflecting student awareness of potential drawbacks. The results further reveal a strong behavioural pathway in which frequent use intention promotes information verification, which fosters responsible use and ultimately improves EWP. This demonstrates that learning outcomes depend on critical, ethical engagement rather than usage frequency alone. Higher education institutions (HEI) should embed AI literacy, verification practices and ethical guidelines into curricula to optimise ChatGPT utility and mitigate dependency. Limitations include a cross-sectional design, lack of contextual factors and a localised sample, which future longitudinal and comparative studies should address. This study offers one of the first empirical examinations linking ChatGPT adoption, verification behaviours, responsible use and measurable EWP gains among Vietnamese TNE students. By integrating cognitive, affective, behavioural and ethical dimensions, it extends traditional technology adoption models to the context of generative AI-assisted academic writing.