Explicability or Openness as the Fifth Principle in AI Ethics: A Proposal and Critique

In the development of AI ethics, the establishment of ethical principles is a crucial aspect. Floridi et al. propose the addition of a fifth principle, namely explicability, to the four foundational principles of bioethics. This approach effectively bridges traditional bioethical principles and the domain of AI, and to some extent, addresses the evolving demands of AI ethics. However, the inclusion of this principle has faced criticism. Cortese and his co-authors argue that explicability is more closely aligned with epistemological ethical rules and should not be classified as a distinct fifth fundamental principle. In this paper, I support this perspective and further propose that explicability be integrated into another ethical principle, openness, thereby strengthening the comprehensive framework of AI ethics.