Generative artificial intelligence is often used as a private assistant for writing, planning, or feedback. Drawing on a sociocultural perspective, this study suggests that AI alone neither ensures optimal learning outcomes nor enables meaningful engagement between learners and AI. In this presentation, we will demonstrate what changes occur when AI is purposefully incorporated as a shared mediating tool within a postgraduate research course. Specific course design incorporating AI tools include utilizing specific AI prompts with weekly learning goals and assessment criteria while the course lecturer modelled how to critically evaluate AI responses, identify limitations, and judge whether feedback was useful. Students then used prompts that they had designed to develop their research ideas, reflected on the AI-generated responses, and brought these reflections to small group discussions, where they compared, challenged, and refined their ideas. At the conclusion of the course, students reported that this process helped them develop more focused research questions, engage more deeply with feedback, build confidence, and interact productively with peers. Our findings offer practical guidance for designing inclusive and responsible AI supported research learning environments in postgraduate and teacher education contexts. They also highlight implications for course design, academic integrity, and institutional AI policy.
Dr Sara Mashayekh is a Lecturer in the School of Education at UNSW Sydney and Program Director of the Master of Education and Master of Educational Leadership. Her teaching and research focus on AI in education, educational research, researcher development and inclusive pedagogy. She is particularly interested in how AI can support learner agency, critical judgement, participation and belonging without replacing human thinking or dialogue.
Jihyun Lee is a Professor in the School of Education at UNSW-Sydney, Australia. Her current research focuses on Generative AI tools—exploring, understanding, applications, and development. Her current GenAI projects include university admission prediction, high school career advising, AI-driven support for mathematics learning, Gen AI tool development aimed at enhancing student well-being, and Gen AI applications for streamlining and supporting student university major selection. At her institution, she contributes to university-wide efforts to develop GenAI tools for all courses and to build Learning Analytics Platforms that monitor student success. Currently, she is a member of the UNSW AI Institute and UNSW Cybersecurity.