Generative AI, Scenario Based Learning, and Educational Communication in Chinese Language Education: A Qualitative Case Study of Vietnamese University Students

Abstract

Background: The rapid rise of Generative Artificial Intelligence is transforming language education by supporting writing, communication, and interactive learning. In Chinese Language Education, integrating AI with scenario-based learning offers new opportunities to enhance intercultural and educational communication, particularly for university students in Vietnam. Aims: This study explores how teacher-mediated generative artificial intelligence (AI) can be integrated into scenario-based Chinese language education for Vietnamese university students, with particular attention to intercultural communication and educational communication. Study Design: The paper adopts a qualitative pedagogical case-study design supported by classroom task analysis, student reception data, teacher evaluation, and reflective evidence. Place and Duration of Study: The case was conducted in an undergraduate Chinese language and culture course at a public university in Hanoi, Vietnam, during one teaching cycle. Methodology: The participants were 36 third-year Vietnamese university students majoring in Chinese Studies. The intervention asked students to use generative AI as a drafting and revision assistant for scenario tasks and multimodal educational communication products. Data included a pre-task learner profile survey, classroom observation notes, AI prompt-and-output logs, student scripts, 52 final video products (17 group videos and 35 individual videos), short reflective comments, and teacher evaluation rubrics. The data were analyzed thematically in four stages: AI-use process, language and intercultural benefits, educational communication outputs, and risks requiring teacher mediation. Results: The findings show that AI-supported preparation reduced the difficulty of beginning Chinese oral and written tasks, expanded students' access to vocabulary and dialogue structures, and helped transform classroom assignments into visible outputs such as role-play videos, bilingual captions, cultural explanation posts, and programme-introduction materials. At the same time, the analysis found over-formal expressions, culturally generic explanations, occasional factual inaccuracies, and weak learner authorship when students relied too heavily on generated text. The strongest learning effects appeared when AI drafts were followed by student revision, peer discussion, teacher correction, and intercultural reflection. Conclusion: Generative AI can support Chinese language education when it is used as a guided assistant rather than a replacement for teachers. In the Vietnamese university context, teacher-mediated AI-supported scenario-based learning can connect language practice, intercultural competence, and educational communication, but it requires transparent AI use, student consent, privacy protection, and teacher supervision to ensure linguistic accuracy, cultural appropriateness, and responsible public-facing outputs.