The Impact of Disclosure: Exploring Public Perceptions of AI-Assisted Journalism in News Reporting

This study investigates the effects of three variations of artificial intelligence (AI) disclosures within a news article in order to determine the effect that disclosures have on readers’ perceptions of the journalistic quality of the article and media trust. A randomized controlled experiment (N = 83) found that, when exposed to an identical news article with one of three versions of a disclosure statement (one containing no mention of AI, one disclosing the use of AI without human supervision, and one containing an AI disclosure with editorial review), readers’ trust in the news organization was found, as hypothesized, to be affected by their perceptions of the journalistic quality of the news article, but those perceptions were not affected by the disclosures themselves and did not cause down-the-line effects on media trust more broadly. Taken together, these results, despite the conventional limitations of relying on a student sample, indicate that the effects of AI on readers’ perceptions of journalistic quality and media trust might be more nuanced than what is commonly discussed in meta-journalistic and public discourse surrounding AI.<p></p>