Speaker
Description
The rapid integration of generative AI tools such as ChatGPT and Grammarly is reshaping EFL translation instruction in Vietnamese higher education, yet empirical evidence from specific institutional contexts remains limited. This study investigates HUFLIT English majors' perceptions of AI-supported translation learning through an integrated Technology Acceptance Model (TAM), Technological Pedagogical Content Knowledge (TPACK), and Self-Regulated Learning (SRL) framework. Using a structured questionnaire administered to 219 second- to fourth-year English majors, the study examined perceived benefits and risks of AI, teacher AI integration, and metacognitive engagement during translation tasks. Descriptive statistics revealed high AI adoption rates (92.7%), strong perceived usefulness and ease of use, and widespread self-reported cognitive overreliance. Exploratory factor analysis (KMO = 0.899; four factors explaining 73.53% of variance) identified a four-factor model: AI Learning Engagement, Teacher AI Integration, Cognitive Overreliance, and Perceived Ease of Use. Notably, perceived usefulness and self-regulated learning converged into a single construct, suggesting that HUFLIT students perceive AI as valuable precisely because it scaffolds their metacognitive processes. These findings underscore the critical role of teacher mediation and institutional policy in supporting responsible, critically engaged AI use in Vietnamese EFL translation programs.
