Abstract
With the rise of generative AIs such as ChatGPT, fake news is easier to create than ever before. As a result, detecting and flagging fake news has become a critical task for social media platforms to ensure a safe environment for advertisers. Few studies have examined how fake news labeling influences the evaluation of a brand whose ad is inadvertently placed alongside fake news. We conducted three experiments using different types of fake news articles and ad stimuli to test fact-checking labeling effects. Study 1 employed a 2 (labeling vs. no labeling) x 2 (content-ad association high vs. low) between-subjects factorial design with a fake news story, disguised as real news, related to a product recall. The study findings revealed that a news story labeled as false hurts the evaluations of the adjoining ad and its brand by undermining the credibility of the story. This credibility spillover effect is moderated by the level of association between the story and the ad. Study 2 found that the same ads used in Study 1 received significantly more positive ad evaluations when they were presented in a positive non-fake news article. To determine whether the effects of labeling would be replicated in non-negative fake news, Study 3 was conducted. Even in non-negative fake news, when fact-check labeling was presented, the placed ad and its brand received significantly negative evaluations.
Original language | English (US) |
---|---|
Journal | Social Science Journal |
DOIs | |
State | Accepted/In press - 2023 |
Keywords
- Brand safety
- Explicit warning
- Fact-checking
- Fake news
- Labeling effect
- Spillover effect
ASJC Scopus subject areas
- Social Psychology
- Sociology and Political Science