Community-in-the-Loop Governance for AI-Enabled Learning Analytics in European Schools - A Socio-Technical Framework from the Schoolwise Project

AI-enabled Learning Analytics systems are increasingly used in schools to transform educational data into indicators, visualizations and decision-support insights for teaching, learning and institutional improvement. Their adoption raises significant socio-technical governance challenges, including opacity, automation bias, metric-driven reductionism, data protection risks, weak contestability and limited participation of the communities affected by data-driven educational decisions. European regulatory frameworks such as the GDPR and the AI Act provide a crucial normative basis, yet schools often lack operational models for translating legal requirements into everyday educational governance. This paper proposes a community-in-the-loop governance framework for AI-enabled Learning Analytics in primary and secondary education. Drawing on the SCHOOLWISE project, the paper frames Learning Analytics as an AI-enabled socio-technical system whose legitimacy depends on regulatory alignment, pedagogical purpose, participatory oversight, technical safeguards and contestability. The proposed framework extends human-in-the-loop approaches through distributed community participation across the lifecycle of AI-supported educational data practices, from problem definition and data taxonomy design to dashboard interpretation, pedagogical intervention and post-decision review. This paper contributes to the CO-GOV-AI agenda [community-in-the-loop governance of AI-enabled socio-technical systems] by addressing the gaps between governance processes and the communities affected by AI-enabled Learning Analytics in education.