Fida Rachmadiarti, Ika Diyah Candra Arifah, Felix Tan, Arif Hidayat
The rapid integration of Artificial Intelligence (AI) into education has advanced personalized learning, yet concerns remain about its alignment with equitable andragogy. This study addresses the need for empirical evidence on how AI personalization affects student learning while considering ethical, social, and andragogical dimensions often overlooked in tech-focused research. Unlike prior studies emphasizing technical efficiency, this mixed-method research explores both functional and ethical implications of AI-driven personalization in higher education. Using data from 150 university students on an AI-powered learning platform, the study examines how AI personalization influences learning performance, with learning experience as a mediator and AI technology acceptance as a moderator. Partial Least Squares Structural Equation Modelling (PLS-SEM) results show that AI-personalized learning enhances performance primarily through improved learning experiences. Moderation analysis reveals that students’ acceptance of AI significantly affects its effectiveness. Qualitative insights further reveal inequities in access, digital literacy, and andragogical application, raising critical concerns around fairness and critical thinking in AI-based learning environments. The study contributes to the field by integrating ethical and human-centered perspectives into the evaluation of AI in education. It advocates for inclusive, transparent, and andragogy-driven AI design, emphasizing that the transformative potential of AI relies not just on its technical capabilities but on its alignment with educational values. The findings provide both theoretical and practical implications for educators, designers, and policymakers aiming to create equitable digital learning systems. © 2026, Malque Publishing. All rights reserved.
Universitas Negeri Surabaya, Indonesia; University of New South Wales, Australia; Universitas Pendidikan Indonesia, Indonesia
