Abstracting Injustice

In this paper, we explore how the neorepublican concepts of domination and antipower can contribute to the surveillance studies literature and a more democratic and participatory approach to technology development and deployment within the criminal justice system. We frame the neorepublican approach as an alternative to the predominant liberal paradigm, arguing that normative surveillance studies scholarship should emphasize the dominating potential of surveillance practices rather than merely trying to limit actual interference in peoples’ lives. To illustrate, we focus on the use of surveillance technologies that capture images of individuals within the US criminal justice system for recognition and/or identification. Facial or other biometric recognition technologies (FRTs) are increasingly built on artificial intelligence and machine learning algorithms (“AI”). Often seen as a faster, more accurate, and less labor-intensive alternatives to human cognition, AI-powered biometric and facial recognition and other image capture technologies have become widely used within public law enforcement agencies around the world. The deployment of these technologies within the US criminal justice system has produced significant forms of injustice, including faulty identification and the subsequent arrest, detention, and incarceration of innocent individuals. These forms of data injustice are often opaque, hidden behind secretive law enforcement practices or commercial secrecy agreements. We draw from neorepublican conceptions of domination and antipower to frame this legal and technological opacity as an abstraction of injustice. We argue that handing important criminal justice decision-making over to code and algorithms designed, owned, and maintained by private interests exacerbates the potential for the public deployment of unjust systems that subject individuals and communities to unwarranted, arbitrary, and uncontrolled state power. Such government interference represents clear forms of data injustice and domination.