The Moral Mirror: LLM-Assisted Reflection on Trolley Dilemmas for AI Ethics Education

초록

This paper investigates how interacting with ChatGPT on trolley dilemmas affects students' reflective thinking. Thirty-seven undergraduates engaged with five trolley dilemma scenarios, comparing their decisions with the Al's. Findings indicate that ChatGPT acted not as an oracle, but as a pedagogical catalyst: validating students' views when aligned and forcing articulation of principles when opposed. Leveraging these insights, we propose the "Moral Mirror,"a structured interactive learning environment designed to foster critical ethical reflection through juxtaposed AI perspectives. © 2026 Copyright held by the owner/author(s).

키워드