생성형 인공지능의 거짓 정보에 대한 신뢰와 확산의도에 미치는 영향 = Trust in and Diffusion Intention of Generative Artificial Intelligence–Based Misinformation

다국어 초록 (Multilingual Abstract)

This study recognizes the increasing risk of misinformation (fake news) diffusion driven by the rapid advancement of generative artificial intelligence (AI) technologies and aims to establish a foundation for preventing the spread of such misinformation. Specifically, it empirically examines the structural relationships among source credibility, digital information literacy, and product familiarity, and their effects on trust in misinformation and word-of-mouth (WOM) intention. A survey was conducted using a scenario in which generative AI provided misinformation about a fictitious “antibiotic coffee” product. The collected data were analyzed using structural equation modeling (SEM) to test the causal relationships among the variables. The results indicate that three out of four hypotheses were supported. The key findings are as follows. First, source credibility exerted the strongest positive effect on trust in generative AI–provided information. This suggests that when misinformation impersonates authoritative or professional institutions, information users are at high risk of accepting it uncritically. Second, digital information literacy had a significant negative effect on trust in generative AI information, demonstrating that higher literacy activates critical thinking and functions as a defensive mechanism by reducing blind trust in online information. Third, trust in information showed a very strong positive effect on WOM intention, indicating that when consumers fail to recognize misinformation and perceive it as credible, they are highly likely to disseminate it to others. Contrary to expectations, product familiarity had a positive effect on information trust (rejected hypothesis), which is attributed to the unique characteristics of the fictitious product used in the scenario. This study empirically elucidates the psychological pathways through which misinformation spreads in a generative AI environment and highlights the manipulation of information sources as the most critical risk factor in the diffusion of AI-generated misinformation.