Exploring Public Attitudes Toward Generative AI for News Across Four Countries

Abstract

As generative artificial intelligence increasingly permeates most life domains, studying how the public perceives, uses, and understands AI-driven tools becomes crucial. Especially relying on generative AI for news seeking, information acquisition, and political opinion formation warrants attention from a democratic point of view. Therefore, we conduct a standardized survey of public assessments toward GenAI for news-related purposes, its adoption by news organizations, and how these attitudes relate to a series of individual-level characteristics across four countries (CH, DE, JP, US). Our findings indicate that audiences do not (yet) extensively use GenAI for news-related purposes and possess relatively limited AI knowledge despite acknowledging various risks and benefits. Cross-nationally, trust in the journalistic deployment of AI-powered tools is relatively low. However, assessments vary depending on individual-level characteristics and macro-level contexts. We conclude by discussing our findings' implications for building and maintaining trust between journalism and its audience.