As background, while AI-generated video technology has advanced rapidly in recent years, various opinions regarding AI technology, including expectations and concerns, can be seen on social media and elsewhere. The purpose of this analysis is to classify people's reactions to AI-generated videos and clarify the trends. The analysis focused on 899 comments out of 1558 comments on one news video about AI videos, excluding comments of 30 characters or less and spam comments unrelated to the video's content. JMP was used for the analysis, performing morphological analysis on the comments, classifying frequently occurring words into 12 categories, and using these as dummy variables. A latent class model was then applied and visualized. As a result, four classes were extracted based on AIC: "Doubts about AI generation" (34.72%), "Concerns about the impact of AI on humans" (26.54%), "Advocacy for legal regulation" (20.40%), and "Pessimistic opinions" (18.34%). Based on the above, by clarifying diverse reaction patterns to AI-generated content, this study provides practical insights that can be used by companies and content creators to understand the risk of online backlash and to consider methods of information dissemination when distributing AI-generated content.
