Predicting AI News Credibility: Communicative or Social Capital or Both?

Abstract

News credibility as an essential democratic value has been at the forefront of scholarly endeavors over the last several decades. Despite prolific research in this area, scholarship on the credibility of algorithm-based and automated news has yet to offer empirical findings in regards to the causes and their impacts. In line with prior studies concerning news credibility, this study examines the driving forces in predicting the level of credibility on artificial intelligence (AI) news. Specially, this study unveils the effects of communicative capital, such as media use and public discussion, among audiences, as well as social capital, such as social trust, on AI news credibility. Data collected through a nationwide online survey reveals that media use through television, social network sites, and online news sites, as well as public discussion yielded a positive association with AI news credibility. Of particular interest is that social trust moderated the effect of public discussion on credibility, indicating that the relationship between discussion and credibility was even stronger for those who have a higher level of trust in others. Implications are further discussed.