How do users respond to AI fact-checkers?

Abstract

Introduction. This paper aims to empirically validate a conceptual model that explains how users respond to AI fact-checkers originating from different countries. Guided by the country-of-origin effect, source credibility theory, and elaboration likelihood model, the model comprises five variables, namely, AI fact-checkers, fact-checker source credibility, perceived credibility of flagged news, issue involvement, and AI literacy. Method. An online experiment was conducted to examine how participants responded to AI fact-checkers from two countries, namely the United States of America (U.S.) and China. Analysis. A total of 139 responses were collected in this study. Data was analysed using a one-way analysis of variance (ANOVA), PROCESS Model 4 and 9. Results. The results showed that AI fact-checkers (country of origin: U.S. vs. China) directly influenced the perceived credibility of flagged news. Fact-checker source credibility mediated the effects of AI fact-checkers. Issue involvement moderated the indirect effect of AI fact-checkers on perceived credibility of flagged news via fact-checker source credibility, whereas AI literacy did not. Conclusion(s). Theoretically, this paper adds to the scholarly understanding of the effectiveness of AI fact-checkers from different countries of origin. Practically, it highlights the importance of considering the country-of-origin effect when deploying AI fact-checkers for social media platforms.