La predictive policing vista dalla prospettiva dei surveillance studies. Un’analisi attraverso i concetti di sorveglianza e dataveillance

Today, predictive policing is undergoing a period of considerable geographical expansion, and its technical evolution is experiencing a season of remarkable growth thanks to the continuous development of algorithmic systems. In scientific literature, predictive policing operations and their consequences are often examined from a procedural perspective, but they also involve a series of processes and elements which occupy a central position in a different, albeit related, scientific field, that is the field of surveillance studies. For this reason, the aim of this paper is to propose an analysis of predictive policing using some conceptual tools specific to surveillance studies. To do so, it seemed appropriate to draw on two central concepts in this field: dataveillance, coined in 1988 by Roger Clarke, and surveillance, which dates back to modern times and is now embodied above all by the phenomenon of video-surveillance. This analysis revealed the composite nature of predictive policing. In fact, it consists of two main phases. The first one involves the collection and storage of data and can therefore be framed within the conceptual boundaries of dataveillance. The second phase involves monitoring by law enforcement agencies of individuals and social groups identified through data: here, we can see the typical features of traditional surveillance. The link between these two phases is an essential moment of algorithmic analysis. The usefulness of adopting an approach based on surveillance studies can thus be twofold. On the one hand, it allows to grasp in a linear manner the typical features of the functioning of predictive policing mechanisms, i.e., a unitary phenomenon resulting from the harmonious intertwining of several distinct phases. On the other hand, the perspective of surveillance studies allows to bear in mind that each of the phases of predictive policing encompasses issues of significance for both the individual and society. In the context of predictive policing, these critical issues do not seem to cancel each other out or mitigate each other. On the contrary, they persist and tend to intertwine, sometimes ending up reinforcing each other. An example of this danger can be seen in the possibility of a “targeted” chilling effect, i.e. an effect typical of surveillance mechanisms that is intended to be produced on individuals and social groups previously identified through dataveillance tools.