Believing the news, contradicting stereotypes? An experiment on the relative credibility of human and AI fact-checkers

This study examines how corruption stereotypes, expectancy violations and fact-checker source influence trust in a fact-check. This research focuses on university students, who exhibit high baseline trust in AI technologies, but remain underexamined in research on trust in AI fact checking. Extending the literature, we show that fact-checks attributed to an AI are trusted less than identical fact-checks attributed to a human fact-checker, even among this AI-literate social-media generation. As the effectiveness of fact-checkers is mediated by source credibility, low trust limits AI’s potential to combat misinformation.