AI-supported adaptive learning in higher education: a multidisciplinary case study on teacher-led personalization, SDG-focused circular economy content, and research methodology development

This study presents the outcomes of an educational action research project that explored the implementation of an AI-supported adaptive learning environment in higher education. The primary objective was to investigate how teacher-driven decision-making processes could leverage AI tools to foster self-regulated learning, enhance group dynamics, and personalize learning pathways in complex project-based courses. The pedagogical intervention integrated research methodology and sustainability topics, with a specific emphasis on circular economy models, thus providing a rich multidimensional context that combined digital pedagogy, cognitive psychology, and environmental education.Over a four-month blended learning period, students engaged in interdisciplinary teams, applying methods of peer assessment, self-reflection, and structured teacher feedback. Rather than automating evaluations, the AI component was designed to support teachers’ interpretative and mentoring work by visualizing discrepancies in learning activity, offering early warnings on motivational drops, and generating tailored feedback suggestions. This approach aligns with emerging literature that conceptualizes AI not as a replacement for pedagogical expertise, but as an extension of teachers’ reflective and strategic capacities within dynamic learning ecosystems.Thematic analysis revealed several significant developments. Students increasingly acted as active designers of their own learning processes, demonstrating higher levels of metacognitive awareness and deliberate planning. Motivation gaps and so-called “free rider” phenomena often turned out to be rooted in mismatches between individual cognitive styles and group progression rates rather than mere disengagement. The AI-enhanced environment enabled timely, differentiated interventions, supporting remediation for slower learners and advanced challenges for those progressing faster. As a result, instructors transitioned from content transmitters to strategic mentors and ecosystem architects who could integrate complex sustainability issues with personalized learning strategies.This research contributes to the broader discourse on how AI-assisted educational models can be leveraged to build more responsive, personalized, and interdisciplinary learning environments in higher education. It offers insights into connecting adaptive learning technologies with pressing global themes such as sustainability and circular economy, and highlights the transformative potential of combining technological decision-support systems with human pedagogical judgment.