Custom chatbots for collaborative learning in higher education

Abstract

Existing research shows that most educational AI applications continue to mirror dominant teacher-centred approaches and primarily support individual and closed learning formats (Xu and Ouyang 2022). As a result, the role of chatbots in more open, collaborative environments such as problem-based learning (PBL) is still insufficiently explored. Recent reviews highlight a persistent misalignment between learning strategies and the ways chatbots are designed and implemented, pointing to issues of limited contextual precision, reduced transparency, and the opaqueness of proprietary systems (Wollny et al. 2021; Hwang and Chang 2023; Labadze et al. 2023; Otto et al., 2025). These challenges become particularly visible in our local educational context of problem-based learning, an educational approach that reflects key elements of networked learning such as collaborative inquiry, shared construction and distribution of knowledge, and work with situated and open-ended problems in project groups. Within project groups, general-purpose chatbots often fall short in supporting the situated nature of students’ work. Their reliance on broad, general knowledge sources can lead to inaccuracies or irrelevant output when students engage with specialised or emergent project contexts. Moreover, the undisclosed or uneven use of GenAI by individual group members can create opacity in collaborative processes, create shadow workspaces and undermine trust and shared responsibility (Otto et al., 2025). The “black box” character of many commercially developed chatbots further complicates efforts to ensure equitable participation and mutual understanding in group work (Bozkurt et al. 2023).

Ähnliche Arbeiten