Machine Teachers in the Classroom: How Self-Efficacy Shapes Student Perceptions of Human-AI Collaboration in AI-Based Education

With the rapid advancement of artificial intelligence (AI), the educational landscape is evolving, increasingly integrating AI into classrooms. In response to this trend, the present study investigates how students perceive human-AI collaboration in AI-based education. Using an online experiment where undergraduate students in the U.S. watched to a lecture delivered by an AI instructor, data were collected to explore these dynamics. The findings reveal two primary insights. First, the impact of human-AI collaboration on students' perceptions varies by their level of self-efficacy in learning. Students with high self-efficacy hold more favorable views toward an AI instructor and AI-based education, particularly when AI-based education is facilitated by human-AI collaboration. Second, the perceived credibility of the AI instructor plays a critical mediating role in fostering positive perceptions of AI-based education. These findings offer significant implications for both research and practical applications in AI in education.