#577: Student, Hybrid, or AI?: Examining Authorial Roles and Network Structures in GenAI-Supported Collaborative Learning

As generative AI (GenAI) reshapes collaborative learning, understanding how authorial roles interact within evolving network structures is crucial. This exploratory case study explored authorial roles and social network structures in a Knowledge Forum-based computational thinking course. We analyzed 527 discussion notes from 54 undergraduates using Social Network Analysis (SNA) and deep learning classification. Notes were categorized into four roles: pure student (Stu), direct GenAI outputs, Student-GenAI (SG) hybrids, and Student-GenAI-Student (SGS) iterations. Results showed student-only contributions dominated (73.8%), with GenAI-involved notes at 26.2%. Network analysis revealed Stu roles as primary information hubs with high centrality, while hybrid SG/SGS roles were less central than expected. Findings suggest that without explicit guidance, students remain core connectors and AI-facilitated knowledge brokerage is underutilized. Educational practice must intentionally scaffold hybrid roles to foster balanced, integrated knowledge flow in GenAI-supported collaborative learning.