The Digital Construction of Artificial Intelligence as a Crime Threat: YouTube Comments within the Framework of Moral Panic

This study aims to reveal how artificial intelligence is constructed as a societal threat, how fear discourse is produced, and how this situation spreads in digital environments. Accordingly, the comments made on the YouTube video titled “Can Artificial Intelligence Commit Crimes?” were analyzed within the framework of moral panic. The study examines which elements of moral panic are reproduced in audience comments.
Out of 229 comments, 56 consisted of feedback about the channel, video requests, or only emojis. Therefore, these 56 comments were excluded, and the analysis was conducted on the remaining 173 comments. The comments were coded according to the five stages of Moral Panic Theory: (1) threat perception, (2) folk devil, (3) concern, (4) solution development, and (5) subsidience. Furthermore, sub-dimensions were created under each category for a more detailed analysis. Folk devil (n=45) and subsidence (n=50) were the most frequently observed categories. In 29% of the comments, artificial intelligence is perceived as a threat (n=30). It is observed that future-oriented anxiety (n=12) and dystopian framing (n=12) emerged at equal levels. In the comments, those most frequently held responsible for AI committing crimes are programmers, designers, developers, users, and technology companies (n=33). The most common solution proposed to prevent AI from committing crimes is the restriction of artificial intelligence (n=16). However, it is also seen that there is a significant subsidience discourse opposing the panic discourse (n=50). Yet, this attenuation is constructed more through humor (n=28) rather than rationalization (n=3). In conclusion, the study shows that moral panic dynamics are reproduced in the YouTube comment section; however, these processes are not one-dimensional and homogeneous, but rather are shaped in a conflictual and multilayered way in digital environments.