The Paradoxes of Artificial Intelligence in Management Education: What We Learn from Providing Human and AI Feedback to Student Global Virtual Teams

This study investigates paradoxes presented by artificial intelligence (AI) in management education as revealed by the experience of a large-scale exercise in which students work in global virtual teams (GVTs). While the role of AI is growing exponentially in most fields of human experience, with a simultaneous increase in demand for students to develop competencies in using AI, the role of artificial intelligence (AI) in management education remains controversial and unevenly applied. Often, AI is partially or completely forbidden to students because it is considered unethical or harmful to creativity and learning. Its use by academics is stigmatized for similar reasons. This situation creates an increasing tension between AI's growing potential and the limitations and hesitations associated with its adoption. In this study, we leverage a cutting-edge global experiential learning project with students working in culturally diverse GVTs (the X-Culture Project) to examine evidence of the paradoxical tensions produced by AI when used to provide feedback to students on their work. We employ an experimental design to compare the satisfaction, performance and creativity of students randomly assigned to one of four experimental conditions: 1) feedback received from humans, 2) feedback received from AI with students told the feedback was generated by AI, 3) feedback received from AI but students are not told that AI created the feedback, and 4) no feedback provided (control group). Our findings deepen the paradoxes surrounding the adoption of AI. There is no difference in student satisfaction between teams receiving human and AI feedback. Also, there is no difference in student satisfaction regardless of whether students know feedback is from AI. Teams receiving AI feedback underperformed compared to those receiving human feedback, but whether team members know feedback is from AI does not make a difference in team performance. Teams receiving AI feedback were also less creative than those receiving human feedback, but if students are not aware of the source of feedback, teams are more creative than if they are. These findings magnify the paradoxes of AI in education by showing both the value and the hazards of using AI to provide feedback on student work.