Integrating AI-Based Machine Translation in Translation Pedagogy: Evidence from a Mixed-Methods Study in Vietnam

Abstract

The rise of artificial intelligence (AI), neural machine translation (NMT), and large language models (LLMs) such as GPT, Gemini, and Claude has transformed translation pedagogy worldwide. However, empirical research on how AI affects translator training in Vietnam remains scarce.

This study adopts a mixed-methods design to examine AI adoption in translation classrooms, combining survey data from 93 students and five lecturers with semi-structured interviews. Quantitative data were analyzed through descriptive and comparative statistics, while qualitative findings were thematically coded.

Results indicate high readiness and positive attitudes among both students and teachers, yet AI applications remain fragmented and unsystematic. Tools such as ChatGPT, Google Translate, and DeepL enhance translation speed, vocabulary, and critical reflection but also pose risks of over-reliance, diminished linguistic sensitivity, and ethical issues. Four key determinants-digital competence, infrastructure, institutional policy, and academic integrity culture-shape effective implementation.

The study recommends a Pedagogically Guided AI Integration (PGAI) model to ensure responsible, sustainable adoption. It advances the concept of AI Literacy for Translators and contributes to the emerging framework of Translation Pedagogy 4.0 in the context of digital transformation.