번역에서 AI 피드백에 대한 대학생들의 인식과 활용

College Students’ Perceptions and Use of AI Feedback in Translation

This study examines how college students with different English proficiency levels perceive and use AI-based feedback in translation. Thirty-three first-year students participated in the study which included surveys, translation tasks, AI usage reports, and interviews. Results show that students widely relied on AI tools such as ChatGPT, Naver, and Gemini, with lower English proficiency learners heavily depending on AI throughout the translation process. Higher English proficiency learners showed the tendency to review and revise AI-generated feedback critically, whereas lower English proficiency learners were more likely to accept it as it was, without any modification. AI tools also showed clear limitations in translating culturally specific expressions, neologisms, and literary texts, frequently producing inaccurate or contextually inappropriate results. These findings reveal that the use of AI underscores the need for guidance to help students recognize the limitations of different AI tools, develop the ability to critically evaluate their work, and refine AI-generated translations.