We are entering an era of AI-Mediated Communication (AI-MC) where interpersonal communication is not only mediated by technology, but is increasingly optimized, augmented, or generated by artificial intelligence with the potential to disrupt existing patterns of communication and online self-presentation.
This study takes a first look at the impact of AI-MC on online self-presentation by examining how people react to Airbnb host profiles that, they are told, may have been generated by AI. We conduct two online experiments to examine how perceptions of trustworthiness of a host represented by a text profile change depending on whether the profile is perceived as AI-generated.
We show, surprisingly, that people do not change their trustworthiness ratings when they are told all profiles were generated by AI. On the other hand, when people are uncertain whether the profile was AI-generated or not, they give lower trust ratings to profiles they perceive as AI-generated. We draw on relevant theories that may explain these findings and offer directions for future AI-MC work.