Calibrating trust in communicative AI Navigating the myths, merits and menaces

More and more people are integrating Conversational Artificial Intelligence into their daily lives, to recommend information, as productivity tools for work or school, or even for emotional support. While useful, these agents also come with several risks, as they can often provide inaccurate or biased answers, and individuals can overestimate their capabilities. In his inaugural lecture, Araujo begins by reflecting on how the relationship between media, organizations, and society has been profoundly changed by technology in recent years. He further discusses what we know about how individuals interact with conversational AI agents like ChatGPT, and what research is needed to develop strategies to enable citizens to adequately calibrate their trust in this technology.