On the moral status of AI entities and robots: a critique of the social-relational approach and a defense of the properties-based approach

Abstract

This chapter critiques the social-relational approach to moral status, specifically in the context of artificial intelligence and robotics. It argues against assessing the moral status of entities like social robots solely based on the nature of their social relationships with humans. Instead, it advocates for a properties-based approach, emphasizing inherent cognitive qualities such as rationality, intelligence, autonomy, and self-awareness. The critique is structured around five principal objections to the social-relational approach: its dependency on human recognition, neglect of intrinsic properties, anthropocentric bias, exclusion of nonsocial AI, and potential for exploitation. The chapter also includes a response to Moosavi's discussion of the paper clip-maximizer thought experiment and its relevance to the issue of moral status. The chapter concludes by recommending a shift towards a framework that recognizes entities' moral status based on species-independent, morally relevant criteria, suggesting that this approach is more inclusive and reliable for integrating intelligent machines into our moral community.