This study examines how the integration of artificial intelligence (AI) into journalism shapes public perceptions of news credibility. It examines differences in credibility toward news authored by AI, human journalists and human–AI collaborations. It also investigates whether AI-related public discussion serves as a mediating mechanism and social trust as a moderating factor.
Drawing on communicative and social capital theories, the study tests a conceptual model linking AI-related news consumption, public discussion and social trust to perceived news credibility across three authorship types. Using data from a nationally representative US panel (N = 1,252), structural equation modeling was conducted to assess both direct and moderated relationships.
News co-produced by AI and human journalists is perceived as more credible than AI-generated news and comparably credible to human-written news. AI-related public discussion serves as a key mechanism that enhances the perceived credibility of AI-generated content. Crucially, social trust moderates these relationships: at higher levels of trust, the positive association between discussion and credibility is stronger across all authorship types, especially for AI-only and collaborative news.
Using data collected in April–May 2022, prior to the public release of ChatGPT, this study offers a pre-generative-AI baseline for understanding audience responses to AI-generated news. It advances a socio-technical framework that integrates public discussion (communication) and social trust (social capital) to explain credibility perceptions across AI-, human- and human–AI-authored news. The findings highlight the roles of human–AI collaboration and community-level trust in sustaining credible and socially accepted uses of AI in journalism.
