Key points are not available for this paper at this time.
This systematic literature review explores the role of Generative AI (GenAI) in facilitating the knowledge construction of learners through discourse within educational settings. By analyzing 28 studies, selected through the PRISMA protocol, the study addresses four primary research questions concerning (1) research context, (2) the learning context of GenAI implementation, (3) the emergent processes of knowledge construction supported by GenAI, and (4) the impacts of GenAI on knowledge construction, including the role of GenAI and influencing factors. The results revealed that GenAI tools, particularly GPT-based models, have predominantly been implemented for both individual and collaborative learning tasks in higher education and K-12 settings, with applications centering on STEM and language learning. The results indicated that GenAI tools may aid in facilitating early phases of knowledge construction (initiation and exploration). However, fewer studies have addressed the potential of GenAI in promoting more complex cognitive phases, such as negotiation and synthesis. The roles identified for GenAI as collaborator, facilitator, and assistant underscore its diverse capabilities within educational contexts, with its impact shaped by such factors as tool-related factors (such as prompt design), pedagogical integration (such as guided instruction), and learner characteristics (such as trust in AI-generated content). The present study conducts an exploratory investigation into the potential of GenAI as a meaningful collaborator in educational discourse, offering comprehensive insights into the effective integration of GenAI to promote deep and sustained knowledge construction.
Jung et al. (Mon,) studied this question.