From Guru to GPT: Reimagining the Mentor–Protégé Relationship in the Age of AI-Driven Digital Learning

Abstract

The emergence of artificial intelligence in education has fundamentally altered the dynamics of the mentor–protégé relationship, one of the oldest and most formative bonds in human learning. This paper examines how AI-driven platforms such as intelligent tutoring systems and large language models simulate, and in some respects challenge, the functions traditionally performed by human mentors. Drawing on three theoretical frameworks—Human–Machine Interaction Theory, Paul Ricoeur’s Narrative Identity Theory, and the Mentor–Protégé Relational Framework—the study investigates how learner identity is reconstructed in the absence of a human guide and what ethical implications arise when algorithmic systems assume the mentor’s role. Analysis of learner experiences reveals four key patterns: initial resistance to AI mentorship followed by measurable intellectual growth; the development of stronger self-directed learning identities; the irreplaceable deficit of empathy, lived experience, and moral imagination in AI systems; and the superior outcomes produced by hybrid models that combine human mentorship with AI precision. The paper further explores the ethical dimensions of AI mentorship, including the commodification of the learner’s vulnerability, algorithmic bias rooted in Western-centric datasets, and the risks to learner data privacy. While AI holds genuine promise as an equalizer in global education extending quality guidance to learners in underserved communities this potential is contingent on inclusive design, cultural sensitivity, and strong ethical governance. The paper concludes that the human core of mentorship remains irreplaceable and advocates for a thoughtfully designed hybrid model as the most educationally sound and ethically responsible path forward.