Collective Sensemaking in Collaborative GenAI Settings : Exploring How Students Interpret and Negotiate GenAI Outputs in Group Work

Sammanfattning

The increasing integration of generative artificial intelligence in education raises important questions about how students interpret, evaluate, and collaboratively work with AI-generated outputs. Drawing on Weick’s sensemaking theory, this study explores how university students make sense of generative AI outputs during collaborative group tasks. Using an interpretive qualitative case study approach, 13 students participated in workshops involving text and image generation tasks with ChatGPT and Google Gemini. Data were analyzed using reflexive thematic analysis, resulting in five interconnected themes. Findings show that collaborative sensemaking with GenAI is highly social, shaped by group dynamics and prior AI experience, unevenly distributed, and primarily guided by plausibility rather than factual verification. While GenAI supported ideation and exploration, it simultaneously constrained imagination and shaped what participants considered feasible or worth pursuing. The study contributes to HCI and educational AI research by extending sensemaking theory into collaborative human-AI contexts, highlighting how generative AI systems influence not only task outcomes but also the social and cognitive conditions under which collaborative interpretation occurs.