Kesan identiti pelaku (manusia atau AI) dan sikap terhadap AI dalam psikologi atribusi moral dalam kalangan pelajar prasiswazah UKM

Abstract

AI is an important asset in the transition of the digitalization era. However, AI raises various ethical concerns that have gained attention recently. Understanding society's perceptions and expectations regarding the moral behavior of AI systems is important to develop ethical guidelines that align with society's values and norms. Therefore, a within-group experimental study on the effect of perpetrator identity and attitude towards AI on moral attribution was carried out on 50 UKM undergraduate students. Data related to attitudes towards AI was collected using a questionnaire based on structure while moral attribution data was collected using 7 moral attribution questions after subjects read 10 scenarios of human misconduct and 10 scenarios of AI misconduct. One-way ANOVA, Pearson correlation, and discriminant analysis were performed using SPSS. The results show that there is a significant difference in moral attribution between human perpetrators and AI perpetrators where subjects attribute moral wrongness, responsibility, awareness, intentionality, and blame. While AI's misconduct is considered more justifiable and has a higher permissibility compared to human misconduct. The findings of the study do not show a relationship between attitudes toward AI and moral attribution regardless of whether it is an AI perpetrator or a human perpetrator. Moral attribution of awareness, responsibility, intentionality, and blame are the main factors that influence the difference in human-AI moral attribution. The findings of this study shed insights into understanding the need for moral attribution for human actors and AI for legal and ethical guidelines in using AI.