Source Study:
AI-enabled future crime
...the pattern of ratings suggests that delegates were particularly concerned about scalable threats, with crimes involving severe harm to single individuals typically rated lower than (possibly lesser, or ill-defined) harms to large numbers of victims, whole social classes or society at large. Group discussions and the rankings that emerged from them were clearly shaped at least in part by contemporary discourse and current events, including anxieties about electoral interference, the spread of disinformation in a rapidly-changing media landscape, and potential erosion of the norms of Western liberal democracy.
See the full study here: https://crimesciencejournal.biomedcentral.com/articles/10.1186/s40163-020-00123-8