Algorithmic Autonomy or Dependence? A Mixed-Methods Study on AI Personalization and Self-Regulated Learning in Higher Education

As higher education shifts toward self-directed learning, AI-driven personalization tools promise tailored support for students. This qualitative study explores how adaptive AI systems (e.g., Smart Sparrow, AI-enhanced Moodle) affect students’ self-regulated learning (SRL), grounded in Zimmerman’s SRL model and Bandura’s Social Cognitive Theory. A focused group of 40 undergraduates from English and International Relations programs at Nakhchivan State University participated. Through structured interviews and classroom observations, we analyzed students' experiences with AI-supported learning. Findings suggest that while personalized feedback and adaptive pacing can enhance engagement, planning, and motivation, they may also reduce autonomy and initiative. Concerns around overreliance and data privacy emerged. The study highlights the double-edged nature of AI in education and offers actionable insights for fostering learner agency alongside technological support.