AI AS A PEDAGOGICAL COMPANION IN HIGHER EDUCATION: TRANSFORMING TEACHING AND LEARNING PRACTICES IN THE ERA OF GENERATIVE ARTIFICIAL INTELLIGENCE
TL;DR
The research concludes that AI is most productive when positioned as a pedagogical partner that enhances human capabilities, rather than as a substitute for instructors or students.
Abstract
This study aims to analyze how artificial intelligence (AI) functions as a learning companion in transforming learning and teaching practices in a higher education setting. The study employed a qualitative approach using a dual-case study design, conducted in two classes in the Education Study Program at Sangga Buana YPKP University in Bandung. The participants consisted of 50 individuals, including 10 faculty members and 40 students who had direct experience with AI-based learning. Data were collected through semi-structured interviews, non-participant observation of AI-assisted learning activities, and document analysis covering course syllabi, learning materials, assessment guidelines, AI usage policies, and student assignment artifacts. The data were analyzed using reflective thematic analysis, combining deductive and inductive approaches, and were strengthened through source triangulation, participant validation, peer discussion, audit trails, and researcher reflexivity. The results of the study indicate that AI serves as a learning companion that expands access to explanations, provides rapid feedback, supports the exploration of ideas, and facilitates personalized learning. The integration of AI also transforms the role of instructors from knowledge transmitters to facilitators, designers of learning experiences, evaluators of information, and mediators of critical AI use. At the same time, students demonstrate increased opportunities to develop autonomy, self-regulation, and reflection on the learning process. However, the effectiveness of human–AI collaboration is significantly influenced by AI literacy, trust, transparency, information accuracy, academic integrity, data privacy, and human oversight mechanisms. The research concludes that AI is most productive when positioned as a pedagogical partner that enhances human capabilities, rather than as a substitute for instructors or students.