AI 얼굴·생체인식 기술의 개인정보 침해 담론 분석: BERTopic 기반 국내 언론 동향 연구

초록

With the advancement of Artificial Intelligence (AI), facial and biometric recognition technologies are widely expanding across various sectors, including public security, finance, and private services. However, alongside the benefits of these technologies, concerns regarding privacy infringement and the potential for a surveillance society are simultaneously escalating. This study applies BERTopic, a state- of-the-art topic modeling technique, to analyze the discourse surrounding privacy infringements related to AI facial and biometric recognition technologies within Korean media. By collecting news article data and analyzing chronological trends in topics and key keywords, this research aims to identify how the media frames both the benefits and risks of these technologies. The analysis reveals that while industrial applications and benefits were emphasized in the early stages of adoption, the discourse has gradually shifted toward specific risk factors, such as data leakage, deepfake abuse, and the necessity for regulation. By utilizing BERTopic to extract latent social issues from vast amounts of unstructured news data, this study intends to provide foundational data essential for future technology policymaking.

키워드