토픽 모델링을 활용한 인공지능(AI) 윤리 담론 분석과 공학교육의 시사점

Abstract

With the rapid spread of generative AI since ChatGPT, ethical concerns have expanded beyond technical issues to include social responsibility, governance, and education. This study investigates the trends, main keywords and topics of artificial intelligence (AI) ethics discourse in South Korean media and explores their implications for engineering education. Using 19,322 news articles on AI ethics published between November 2022 and May 2025, this study applied text mining techniques including TF-IDF, N-gram, and Latent Dirichlet Allocation (LDA) topic modeling to identify key themes. Results show a steady rise in media coverage, with core keywords such as education, ethics, digital, responsibility, and ESG management. The analyses revealed that discussions of AI ethics are structured around interconnected domains of industry, education, and governance. In particular, AI ethics education and digital competency emerged as a central topic, closely related to industrial innovation and ethical management. Based on these findings, this study offers implications for engineering education across social, educational, industrial, and policy domains, underscoring the need to integrate AI ethics more systematically into curricula and to prepare engineers capable of addressing the ethical and data-driven challenges of the AI era. The study contributes insights for strengthening ethical literacy and designing responsible AI education in the age of AI transformation.