Artificial Intelligence in University English Classes in Taiwan: Opportunities, Risks, and a Context-Sensitive Framework for Integration.

This review examines the use of artificial intelligence (AI), particularly generative AI, in university English classes in Taiwan, where English as a foreign language (EFL) learning intersects with English-medium instruction (EMI) and the Program on Bilingual Education for Students in College (BEST). The paper synthesizes Taiwan-based studies of ChatGPT-supported EFL writing, international research on generative AI in higher education, and policy guidance from the Ministry of Education and UNESCO. Results: The review identifies key benefits, including individualized feedback, writing-process support, low-stakes language practice, EMI preparation, and reduced routine teacher workload. It also highlights major risks, including overreliance, uneven feedback quality, academic integrity concerns, privacy issues, inequitable access, and the possible weakening of critical thinking. AI should be treated neither as a simple solution to English proficiency challenges nor as a threat to be excluded from classrooms. Instead, it should be integrated through transparent assessment, process-based writing tasks, teacher professional development, and critical AI literacy. The paper proposes a context-sensitive framework for AI integration in Taiwanese higher education and calls for future research that examines durable language learning rather than short-term performance gains.