Algorithmic Transformations in Policing Practices

Summary

Algorithms are reshaping policing practices, influencing how decisions are made, suspicion is formed, and discretion is enacted. Policing is increasingly structured around data-driven systems, facilitated by a range of algorithmic-based technologies such as data-driven dashboards, risk scores, facial recognition, and many more. Algorithmic systems are shifting policing from a reactive to a preemptive model as tools such as risk scoring attempt to predict who could be a victim or a perpetrator of a crime. Administrative and background tasks are being automated with algorithmic systems, from tools that automatically redact documents or complete official forms on behalf of officers, to those that conduct investigatory work on gathered data. In some instances, algorithmic systems, such as drones as first responders or AI chatbots answering nonemergency calls, are replacing human police officers and operators, raising concerns on the deskilling of the police.

Algorithms are therefore transforming policing practices in a variety of different ways by altering how risk is identified, who is monitored, and what is prioritized. Some changes, such as the platformization of policing, are altering the very infrastructure on which policing operates. Together these changes affect police decision-making processes, and as a result, how police officers exert discretion and act on suspicion. These parts of policing are increasingly being delegated to algorithm-based systems that triage police contact and automate suspicion through data analytics. This is changing the patrol functions of the police and how they interact with the general public. Some algorithmic tools may streamline aspects of policing, increasing efficiency and performance, but they are also perpetuating inherent biases and fundamentally reshaping policing practices.