Facing the pain: ethical considerations of AI-based pain detection of farmed animals

Abstract Automated pain detection (APD) is an emerging technology that runs on artificial intelligence (AI) (e.g., machine learning, computer vision, and deep learning) and is aimed at detecting pain in farmed animals. One type of APD analyses the facial expressions of farmed animals, indicated by movements in facial muscles, to identify indicators of pain. This paper evaluates six ethical concerns of APD use in the agri-food sector: (1) harm caused by incorrect diagnosis from APD; (2) harm caused by using drones or robots to collect images; (3) the insufficiency of focusing only on pain for animal welfare; (4) increased alienation within the human-animal relationship due to APDs; (5) the need to focus on tackling already known causes of animal pain; (6) the need to focus on tackling the root cause of farmed animal pain (i.e., animal farming industry). This paper proposes four principles for responsible APD development and use in the agricultural sector: (1) Implement the precautionary principle when there is uncertainty about animal pain; (2) Humans must be kept in the loop on decision-making on animals’ health and well-being; (3) Individuals (e.g., farmers) should be held accountable for animal health or welfare problems that are detected by their APD systems but not investigated or treated; (4) An open-source approach must be established to train APDs with high-quality images, ensuring the technologies work as well as possible.