Human Decision-Making and Editing Behavior in AI-Supported Translation: Evidence from EFL Student Translators

Truong Sa Nguyen, Pham Thien Thu Le, Nhat Sinh Vo, Thi Thuy Le, Ngoc Hiep Nguyen

Many recent studies have investigated different types of technologies and AI-powered tools employed, and perceptions and evaluation of the AI users in translation. This study further closes the gap of literature by examining students’ behavior and decision-making process in different stages of translation tasks. The study was conducted with a group of 33 undergraduate EFL learners in a translation course in which AI tools were encouraged to use by the teacher. Data was collected from participants’ self-reflection survey, and individual stimulated recall interviews. The findings indicate that students’ decisions to employ AI are shaped by multiple interacting factors; rather than relying on AI uncritically, students positioned themselves as primary decision-makers, using AI selectively. Editing behaviors focused mostly on semantic and pragmatic, accompanied by higher-order cognitive processes. The study highlights the central role of humans and suggests that effective integration of AI in translation education requires careful pedagogical practice.