Study of AI-generated assessment in moral decision-making

Abstract

The present study aimed to explore whether and how unfounded AI-generated assessments affect individuals' moral decision-making. A total of 191 Chinese participants were randomly assigned to two groups (AI vs. human) and completed a series of moral dilemma questions and the Oxford Utilitarianism Scale (OUS) on the first day. After receiving groundless assessments with directional implication towards utilitarianism from AI or psychologist on the second day, participants finished another set of moral dilemmas and the OUS again. The results indicated that participants made more deontological choices in hypothetical questions after receiving the groundless AI assessment, while their general moral tendencies remained stable. And the changes were more pronounced in weak moral intuition questions compared to the strong ones. No significant difference was found between the group of AI and the group of psychologist in terms of changes in moral decision-making scores. These findings highlight the importance of considering factors such as emotional arousal, cultural norms, and situational factors when investigating the impact of baseless AI assessments on moral choices. And further research is needed to explore the influence of AI in various cultural contexts and in more subjective evaluation scenarios.