Abstract
This study develops and empirically evaluates the effectiveness of a triadic teaching model - “Instructor-led, AI-supported, Student-autonomous” - implemented with a sample of 242 third-year Chinese language students at Ho Chi Minh City University of Foreign Languages - Information Technology (HUFLIT). Grounded in the theoretical frameworks of cognitive enhancement, desirable difficulty, and digital space separation, the analysis demonstrates that the model yields consistent improvements in academic performance across all participants. Moreover, it functions as an effective cognitive scaffold for students with weaker initial proficiency, supporting differentiated learning outcomes and highlighting its potential to underpin a sustainable, symbiotic learning ecosystem.
