Illusion or Persuasion: A Study on the Factuality of Generative Artificial Intelligence (GAI) Content and Rhetorical Communication Effects

Abstract

(GAI)to investigate the role of rhetorical expressions on the trust of the public towards GAI-generated content, this paper utilizes control-group experiments. It is shown that intemperate rhetorical expressions in GAI-generated content are conducive to increasing the public’s perception of trust towards it and increasing the likelihood of information sharing. At the same time, these intemperate rhetorical expressions can help to ease the credibility crisis caused by inaccuracies contained in the GAI-generated information to a certain extent, especially toward low-competent groups in terms of media literacy. In this way, GAI hallucinations cannot be just seen as a technological error; they are also a communicative effect, and such findings offer empirical evidence and actionable ways of how to counter the misleading mechanisms of AIGC and governance measures.