VietTESOL International Convention 2026

This study examines whether elements of self-directed learning (SDL) emerge in first-year students’ use of artificial intelligence (AI) tools for academic writing. Grounded in the framework of self-regulated learning (SRL), the study categorises AI-assisted writing behaviours into three dimensions: AI-mediated planning regulation, AI-mediated performance regulation, and AI-mediated reflective regulation. It also explores how these behaviours relate to the perceived benefits and negative impacts of AI-assisted writing. A quantitative research design was adopted, and data were collected from 145 first-year university students at Thu Dau Mot University through a structured questionnaire. Partial Least Squares Structural Equation Modelling (PLS-SEM) was employed to analyse the relationships among AI usage frequency, AI-mediated regulatory behaviours, linguistic knowledge, language use, and negative impacts. Reliability, validity, and mediation analyses were conducted to evaluate the proposed model. The findings indicate that frequent AI use is associated with all three forms of AI-mediated regulation. Planning regulation showed a moderate positive relationship with language use and linguistic knowledge, while reflective regulation was modestly associated with linguistic knowledge. Performance regulation demonstrated limited influence on learning outcomes and a slight association with negative impacts. Mediation analysis suggested that planning and reflective regulation partially explained the relationship between AI usage frequency and positive learning outcomes. The study contributes to research on AI-assisted language learning by suggesting that the educational value of AI tools depends not only on frequency of use but also on how students regulate their learning through AI during the writing process.