AI-GENERATED CONTENT AND AUDIENCE TRUST: A COMPARATIVE STUDY OF HUMAN-PRODUCED AND AI-PRODUCED NEWS STORIES

Abstract

The rapid integration of artificial intelligence (AI) into journalism has transformed the news production landscape, raising important questions about how audiences perceive and trust AI-generated content compared to traditional human-written news. This study examines the effect of news authorship on audience trust by comparing human-produced and AI-generated news articles. A quantitative experimental research design was employed, using a between-subjects approach in which 312 digital news consumers were randomly assigned to read either a human-written or AI-generated news article. Following exposure, participants completed a structured questionnaire measuring audience trust using a validated Likert-scale instrument. The findings reveal a statistically significant difference in trust between the two conditions, with human-produced news receiving higher trust ratings than AI-generated news. The results of the independent samples t-test confirmed this difference (t = 5.84, p < 0.001), while effect size analysis indicated a medium-to-large practical significance (Cohen’s d = 0.66). Regression analysis further showed that authorship type significantly predicted audience trust, explaining 18% of the variance. The study concludes that despite advancements in generative AI, human journalism continues to hold a trust advantage among audiences. These findings highlight the importance of authorship cues in shaping credibility perceptions and provide valuable implications for news organizations adopting AI-driven content generation.