VietTESOL International Convention 2026

Drawing on Stoller’s (2006) theoretical foundation for project-based language teaching, Long’s (2015) and Ellis’s (2003) principles of task-based language teaching, and Thomas’s (2017) research on technology-mediated project-based language learning, this presentation argues that academic integrity is best supported through process-oriented project design that makes language development visible throughout the learning cycle in LLM-supported learning environments. Using examples from a semester-long English for Academic Purposes (EAP) research project, the presentation illustrates how large language models (LLMs) can be integrated into project-based learning while maintaining academic integrity and supporting authentic language production. Students progress through topic selection, annotated bibliography writing, survey design, qualitative interviews, data analysis, research report writing, and presentations, creating multiple opportunities for language use, reflection, and assessment throughout the semester. The project identifies stages where LLM use can support learning and stages where students must demonstrate independent understanding and language production through assessment tasks. Rather than positioning AI as a threat to academic integrity, the presentation demonstrates how carefully designed projects can regulate LLM use, guide learner behaviour, and provide evidence of genuine language development. Participants will receive project materials, assessment procedures, and practical recommendations that can be adapted to other L2 learning contexts.