Japanese undergraduates’ use of ChatGPT in an English public speaking course

Abstract

[EN] This exploratory case study investigates Japanese undergraduates’ perceptions of ChatGPT in an English public-speaking course. Over four weeks, C1-level (CEFR) students engaged in parliamentary-style debates, roleplaying Japanese political figures on current social issues, focusing on speech writing and argument construction. Data were collected from reflective journals (n = 16), semi-structured interviews (n = 3), and classroom field notes. Students perceived ChatGPT as a convenient tool for idea generation, structuring arguments, and anticipating counterarguments. Although students reported that using ChatGPT helped them to improve the logical flow of their speeches, its outputs were often perceived as formulaic, lacking originality, rhetorical depth, and emotional appeal. Some valued ChatGPT as a scaffold for organizing arguments, while others worried it encouraged over-reliance and superficial engagement. Ethical concerns arose regarding authorship, academic integrity, and AI’s influence on rhetorical voice. The findings highlight ChatGPT’s contradictory role as both an enabler of rhetorical structuring and an impediment to deep critical engagement, underscoring the need for AI literacy frameworks that promote self-regulated and reflective AI use in public speaking contexts.