AI Ethics Guidelines: Time to Include Animals

Artificial intelligence (AI) ethics guidelines make strong appeals for ensuring that AI benefits human beings without unjustifiably harming them. Yet these guidelines largely overlook AIs ethically relevant positive and negative impacts on sentient nonhuman animals. This paper argues that it is time to include sentient animals in AI ethics guidelines produced by governments, supranational and intergovernmental bodies, NGOs, academics, professional groups, for-profit corporations, and the like. AI ethics guidelines should explicitly and adequately include values and principles that protect and promote sentient animal interests. The case proceeds by considering five possible objections to recognising animals in these guidelines: 1) Animals and their interests are not significantly affected by AI; 2) Humans lack duties to protect or help animals; 3) Considering animals is overridden by more pressing duties to humans; 4) There is too much disagreement about how animals should be treated; 5) AI ethics guidelines are unhelpful or counterproductive. These objections, the paper argues, ultimately fail. A very important reason for being more inclusive of animals, it claims, is their ongoing moral and social invisibility. Paying more attention to sentient animals as morally significant interest-holders in the rapidly evolving AI ethics discourse would go one step towards redressing this problem of marginalisation. Finally, this paper makes broad suggestions for how animals might be appropriately incorporated in ethics guidelines.