생성형 인공지능(AI) 저널리즘 준칙의‘적용 가능성’과 ‘한계’에 관한 연구: 현업 기자 인식을 중심으로

초록

This study empirically examined the applicability and effectiveness of generative AI journalism guidelines in the context of the rapid expansion of generative AI technologies in news production. To achieve this, journalism guidelines were reorganized into four upper-level categories and eleven subcategories, and in-depth interviews were conducted with fourteen working journalists. The findings showed that journalists generally agreed on the necessity of AI journalism guidelines, but perceived their practical applicability in real newsroom environments to be relatively limited. This suggests that while AI journalism guidelines are accepted at a normative level, their implementation is constrained by various practical limitations in newsroom settings. In addition, journalists’ perceptions varied depending on how closely each guideline was connected to traditional journalistic values and on the level of trust in generative AI technologies. Provisions related to fact-checking and error correction showed relatively high acceptance, whereas provisions concerning transparency, copyright, and governance repeatedly revealed concerns regarding ambiguous standards and gaps between guidelines and actual newsroom practices. In contrast, provisions prioritizing human-created content and restricting synthetic representations of real-world materials showed relatively high acceptance due to low trust in AI technologies and concerns about the distortion of reality. Finally, technological factors and guideline-related factors were identified as the most influential variables, while clarification of guideline standards, expansion of case-based education, and establishment of organizational support systems were suggested as major improvement measures.

키워드