Dialogue Game-Based Learning for AI Ethics Education

Abstract

This study aims to develop a dialogue game for exploring ethical issues in artificial intelligence (Al) and to investigate the dialogic learning process in gameplay. The research suggests that effective Al ethics education goes beyond traditional dogmatic instruction, advocating for the integration of dialogue game-based learning as a promising pedagogical approach. Through the card game Al Ethics Dilemmas, participants articulate their ethical perspectives via moral reasoning discourse. The designed game incorporates three components—situations, roles, and questions—to facilitate players' dialogic learning and ethical deliberations. The study involved 28 Korean learners grouped into 4-6 members each, actively engaging in playing the designed game. Gameplay discourse was recorded and transcribed using Alexander's (2020) framework of learning talk. The analysis revealed that learners' in-game discourse predominantly focused on transactional and expressive aspects. Moreover, participants' discussions indicated the ethical reasoning process, reflecting the core values of Al ethics and the conflict structures surrounding diverse stakeholders. This study underscores the potential of dialogue game-based learning in fostering learners' ethical deliberations and enhancing their understanding of complex Al-related dilemmas.