Bearly Recognizable: Facial Recognition and the Wild

Abstract

While big tech companies are growing more circumspect about the use of facial recognition for humans, interest in nonhuman facial recognition is surging. The identification of animals at the level of the individual, rather than the species, has long been of interest for ecologists and others. Drawing on research with scientists and computer programmers working on computer vision tools for ecology, this paper examines the development of a facial recognition tool for bears, highlighting the contingent social, material, and technical processes involved in new practices of digitally monitoring wildlife. Making animal facial recognition entails the creative repurposing of computer vision tools like object detection systems and reflects the social processes that shape the accumulation of data about animal life (and make biometric identification possible). By bringing together science and technology studies (STS) research on algorithms and artificial intelligence with studies of the environment and wildlife, the article also addresses how machinic processes of identification relate to longstanding debates over what the human relationship with wildlife and wilderness should be. Facial recognition tools promise to individualize bears at scale: these new forms of individuation and identification produce new affective between humans and wild animals that unsettle boundaries and categorizations like wild or domestic.