Examining the Role of Genuine Emotions for Trustworthy AI

Trust is a fundamental part of interpersonal relationships and due to the rise of AI systems in many domains, we must examine their trustworthiness. While it is possible to rely on AI to perform certain tasks in certain situations, they do not warrant interpersonal trust. An AI can only be considered trustworthy if it is normatively or affectively motivated to act as expected. We present the argument that the absence of genuine emotions in current AI systems renders them untrustworthy, even if they appear emotional by exhibiting social cues.