Artificial Intelligence in Broadcasting: Public Trust and Misinformation Detection in Robotic News Presentation

Abstract Purpose : Social robots are increasingly proposed for high-stakes societal roles such as journalism. This raises questions about public trust, especially for non-humanoid designs, and how source credibility affects an audience's critical evaluation of information. Methods : In a field experiment, 40 participants in Brazil interacted with an emotionally expressive, non-humanoid social robot that presented three news stories, one of which was fabricated. Pre- and post-interaction questionnaires assessed perceived trust, role acceptance, and misinformation detection. Results : The robot achieved high levels of credibility, and this trust correlated positively with its acceptance in journalistic roles. However, this credibility did not confer the ability to discern the truth; participants failed to identify the fabricated story at a rate below chance. Furthermore, the brief interaction was insufficient to significantly alter the participants' pre-existing concerns or perceived advantages of the technology. Conclusion : The findings reveal a critical 'credibility paradox', where a trusted robotic agent may inadvertently lower an audience's critical scrutiny, increasing vulnerability to misinformation. Public perceptions of robotic journalists appear to be robust and not easily swayed by short-term interactions. This study suggests that designing for credibility alone is insufficient; a parallel focus on fostering critical engagement is essential for the responsible deployment of automated agents in our information ecosystem.