Computer-Assisted Translation and ChatGPT: A Case Study in Teaching English-French Translation at Euromed University of Fès

Abstract

This study investigates the integration of ChatGPT-4, an AI language model, in translation tasks within a classroom environment, focusing specifically on English-French translation activities performed by 13 students. Employing a mixed-methods approach, data were collected through focus groups, classroom observations, reflective journals, and semi-structured questionnaires to gain a comprehensive understanding of students’ perceptions, interaction patterns, and learning strategies when engaging with AI-assisted translation tools. The findings reveal that a majority of students perceived ChatGPT-4 as a useful resource for vocabulary and syntax suggestions, often utilizing it primarily as a reference or drafting tool to support their initial translations. Despite these advantages, participants widely acknowledged the necessity for thorough post-editing, as ChatGPT-4 frequently produced literal or semantically inaccurate translations requiring correction. The use of ChatGPT-4 also encouraged students to develop critical thinking and editing skills by engaging more actively in evaluating and refining AI-generated outputs. While some students initially exhibited skepticism or anxiety about relying on AI, their confidence in using the tool grew with practice, highlighting the emotional and pedagogical challenges involved in adopting new technologies. Concerns about potential overreliance on AI were also expressed, underscoring the need for balanced integration. Overall, the study emphasizes the pedagogical potential of ChatGPT-4 as a complementary tool that supports but does not replace human judgment in translation education. It advocates for guided instruction to enhance AI literacy and maintain translation quality.