News bylines and perceived AI authorship: Effects on source and message credibility

Abstract

With emerging abilities to generate content, artificial intelligence (AI) poses a challenge to identifying authorship of news content. This study focuses on source and message credibility evaluation as AI becomes incorporated into journalistic practices. An experiment (N = 269) explored the effects of news bylines and AI authorship on readers’ perceptions. The findings showed that perceived AI contribution, rather than the labeling of the AI role, predicted readers’ perceptions of the source and the content. When readers thought AI contributed more to a news article, they indicated lower message credibility and source credibility perceptions. Humanness perceptions fully mediated the relationships between perceived AI contribution and perceived message credibility and source credibility. This study yielded theoretical implications for understanding readers’ mental model of machine sourceness and practical implications for newsrooms toward ethical AI in news automation and production.

Keywords

• Artificial intelligence
• News credibility
• Humanness
• Communication technology
• Human-AI interaction