What exactly do we mean when we say that people “trust” news media? How has news media trust been conceptualized and measured over the past seven decades? And as artificial intelligence becomes increasingly involved in news production, do our existing theories of news trust still work in the same way?

This talk approaches these questions through three connected studies. I first draw on a systematic review of 623 empirical publications from 1951 to 2025 to examine how news media trust and related concepts have been conceptualized and operationalized across time and contexts. I then turn to AI-generated and automated news. An online experiment examines whether explaining who develops an AI system and how it produces news can shape audiences’ credibility judgments. Finally, I present a longitudinal study asking whether the reinforcing relationship between news use and trust found in traditional journalism also emerges for automated news. Using three-wave panel data, the study finds an asymmetric pattern.

Together, these studies move from how we define and measure news trust, to how AI-related cues influence credibility, and finally to how trust develops through repeated interaction with automated news. The talk considers what remains transferable from traditional news trust research and what may need to be reconsidered as AI becomes increasingly embedded in journalism.

About the speaker:

LIU Yongliang(leo) is a postdoctoral researcher at the University of Cambridge. He received his PhD in Journalism and Communication from Tsinghua University in 2025. His research focuses on AI in journalism, media trust, and risk communication, with a broader interest in how emerging technologies reshape the production and evaluation of journalism. His work has been published in Communication Research, Journalism & Mass Communication Quarterly, Journalism, and Journal of Broadcasting & Electronic Media.