predictivepolicingwatch.com – Addressing Obstacles to Accountability and Transparency through the Compilation and Publication of Largest Repository of Data on Predictive Policing Use to Date

Abstract

Despite the increase in attention paid to police practices–especially those considered to be potentially illegal or biased–over the past decade, the spread of predictive policing use throughout the United States has gone relatively unnoticed and unregulated. For a technology that raises major questions about the legality of its use in arrests and court cases as well as its potential to contribute to biased policing, that most people associate the term “predictive policing” with the film Minority Report, rather than a real police practice, is as confusing as it is concerning. While there are allegedly 150 cities in the United States that utilize predictive policing technology, there is no known record of which cities those are. Only ten cities in the United States have any laws requiring city councils explicitly notify the public when discussing the adoption of new police technology–about 20 additional cities and two states have currently proposed similar laws. And only one city in the United States–New York City–has any law regulating the use of automated decision systems like predictive policing algorithms, which it only passed in 2018. This thesis asks: What is responsible for the gap between real predictive policing use and public knowledge/government action? In the process of answering this question, this thesis attempts to simultaneously understand that gap by highlighting ways in which standard city and vendor practices hide information on predictive policing use, as well as bridge it by presenting the most complete and detailed account of predictive policing use in the United States to date and publishing it online.