Perceptions, attitudes, and behavioral responses of Arab audiences to AI-labeled news: an experimental study grounded in the Technology Acceptance Model
Menna Elhosary, Rasha AbdullaAbstract
Purpose
Empirical evidence on how news audiences actually perceive and respond to AI-labeled news is still scarce. The present study bridges this research gap by applying the Technology Acceptance Model to the underexplored domain of generative AI in news production. It investigates the key factors that shape audience attitudes toward using generative AI in news production, namely Perceived Usefulness, Perceived Ease of Use, and Generative AI’s Awareness/literacy. It also examines how formed attitudes influence perceptions of trustworthiness and sharing intention of AI-labeled news items.
Design/methodology/approach
The work depends on a randomized between-subjects online experiment with pre-exposure measures and post-exposure outcomes on a purposive sample of Arab social media users ( n = 420).
Findings
The results showed that perceived usefulness and ease of use are key drivers of audience attitudes. However, it remains unclear what shapes these perceptions, as the hypothesised correlation with generative AI awareness was only partially supported. Additionally, while AI labeling was found to enhance trust in news, it does not directly translate to a higher intention to share AI-labeled news items.
Practical implications
We provide actionable insights for news organizations seeking to integrate generative AI tools while maintaining news transparency and audience trust. Furthermore, by uncovering the ethical concerns that shape audiences’ negative attitudes, we pave the way for developing strategies to address these concerns and foster more informed audiences.
Social implications
The study underscores the potential of news transparency through AI labeling in fostering trust yet highlights the complexities surrounding audience willingness to engage with AI-labeled news items, as concerns about misinformation, bias, and threats to news integrity remain significant barriers. These concerns point to a potential skepticism that hinders the broader acceptance of integrating generative AI in news production, which implies that news transparency must be backed up with strict ethical guidelines to foster greater acceptance among news audiences.
Originality/value
The work is among the earliest to apply the Technology Acceptance Model to the underexplored domain of AI-labeled news, especially from a non-Western audience perspective.