This study explores the dynamics of mathematics learning through the integration of Artificial Intelligence (AI) applications within the Teams Game Tournament (TGT) method at SMA Negeri 1 Bangorejo. Using a qualitative case study approach, data were collected through in-depth interviews with two mathematics teachers and five students, supplemented by classroom and environmental observations. The findings reveal four specific learning dynamics: (1) a structured five-stage learning sequence that systematically operationalizes constructivist knowledge construction; (2) a categorical shift in the teacher's role from sole instructor to multi-layer facilitator; (3) student activeness driven by collective accountability to the team rather than individual obligation; and (4) differential AI functions adapted to each student's learning needs. Students' responses demonstrated five distinct profiles ranging from affective transformation and collaborative satisfaction to reflective independence and increased self-confidence. Four benefit dimensions were identified — pedagogical, cognitive, social, and metacognitive — accompanied by four categories of challenges and three adaptive strategy groups developed organically in the field. This research contributes four grounded concepts: Multi-Source ZPD, Collective Accountability, AI Psychological Safe Zone, and Organic Peer Scaffolding (OPS).
Copyrights © 2026