DESIGNING A STUDENT-CENTRIC FRAMEWORK FOR EXPLAINABLE AI IN ADAPTIVE LEARNING SYSTEMS: TOWARDS TRANSPARENT, TRUSTWORTHY, AND ETHICAL AI IN EDUCATION

The integration of Artificial Intelligence (AI) technologies into educational environments—particularly through Adaptive Learning Systems (ALS)—has fundamentally transformed the delivery of instruction and personalized learning. These intelligent systems leverage machine learning algorithms to continuously analyze learner data and adapt content, pacing, and pedagogical strategies to individual student needs, leading to enhanced engagement and improved learning outcomes. However, this rapid digital evolution has surfaced critical concerns about the opacity, fairness, and accountability of AI-driven decisions. Often referred to as “black-box” systems, many AI models offer little to no insight into how they arrive at recommendations or interventions, thereby posing challenges related to transparency, interpretability, trust, and ethical governance. This paper argues that to responsibly integrate AI in education, especially in contexts where student data and futures are at stake, it is essential to move beyond system performance and embrace Explainable AI (XAI)—a design philosophy and technical paradigm focused on making AI decision-making understandable to humans. We propose a student-centric framework for XAI in ALS, which places learners and educators at the center of AI design and deployment processes. The framework is informed by interdisciplinary literature and practical case studies in educational technology and AI ethics. It emphasizes three foundational pillars: (1) Human-centered design, which ensures that AI interfaces and feedback mechanisms are accessible and usable by students and educators alike; (2) Stakeholder collaboration, which fosters ongoing dialogue between developers, teachers, learners, ethicists, and policymakers; and (3) Ethical integration, which embeds principles such as fairness, inclusivity, data privacy, and algorithmic accountability into the development lifecycle of adaptive learning tools. The framework also outlines actionable strategies to ensure that explanations provided by AI systems are contextually meaningful, developmentally appropriate, and pedagogically aligned. It considers the differentiated cognitive needs of various learner groups and promotes educational transparency as a right, not a privilege. Ultimately, this student-centric XAI framework aims to cultivate trustworthy, ethical, and pedagogically sound AI systems that not only support learning but also uphold the rights, agency, and dignity of learners in the digital age. By advocating for transparency and ethical accountability, this work contributes to the growing body of research on responsible AI in education and offers practical guidance for institutions aiming to deploy adaptive technologies in equitable and explainable ways.