Advancing socially shared regulation in collaborative learning with AI

In a rapidly changing world, the key question is which human skills are needed to sustain human learning in the future. An essential competence for contemporary learners is the ability to control, monitor, and adapt their own learning, referred to as self-regulated learning (SRL). SRL is a proactive and cyclical process in which learners engage in planning, task execution, monitoring, and reflection to achieve learning goals. As learning increasingly occurs in collaborative and technology-rich environments, regulatory demands extend beyond the individual to include the coordination of shared cognitive, socioemotional, motivational, and metacognitive processes.

My research shows that effective learning in groups is not just about managing tasks, but about regulating group members’ thinking together, namely socially shared regulation in learning (SSRL). In SSRL, group members jointly negotiate goals, monitor progress, and adapt strategies. However, learners often face difficulties in recognizing regulatory challenges and enacting appropriate regulatory responses during collaboration.

Advances in digital learning environments and learning analytics have enabled the use of multimodal data, including learning traces, discourse, and physiological measures, to study self- and socially shared regulation processes. In my research, we have developed multimodal methodological approaches to make these complex regulatory processes observable and to identify moments of challenge and opportunity for regulatory support.

In this master class, I explain advancements in socially shared regulation theory. I will demonstrate how multimodal data and analytics can be used to detect regulatory challenges and support learning regulation at individual and group levels. I will also present insights from our ongoing Hybrid Intelligence research, which examines how human learners and artificial intelligence (AI) can learn together. Our work emphasizes the use of AI to enhance learners’ agency and regulatory capacity, rather than replacing it. From this perspective, future-ready learners require strong self-regulated and socially shared regulation skills, increasingly augmented by AI.