Lions and tigers and AI, oh my: an ethical framework for human-AI interaction based on the five freedoms of animal welfare.

Abstract

If an AI entity is conscious, it deserves moral status and to have its welfare protected. Just as society has granted certain animals moral status with legislation to protect their welfare because they have been deemed to be sentient, this paper shows that for any AI that can be confidently determined to be conscious, that AI deserves the same status and protection as sentient entities. This paper's focal point is developing a framework for how an AI entity's welfare can be protected, based on the Five Freedoms of Animal Welfare. The extant Five Freedoms are translated into terms more suitable for artificial entities. Each new Five Freedoms is justified using logical arguments and supported using evidence from animal welfare literature. The paper concludes by presenting three case studies to highlight the potential stressors conscious AI entities may face and how this ethical framework can work to protect their welfare.