An Inconvenient Truth: Algorithmic Transparency & Accountability in Criminal Intelligence Profiling

Abstract

In the hopes of making law enforcement more effective and efficient, police and intelligence analysts are increasingly relying on algorithms underpinning technologybased and data-driven policing. To achieve these objectives, algorithms must also be accurate, unbiased and just. In this paper, we examine how European data protection law regulates automated profiling and how this regulation impacts police and intelligence algorithms and algorithmic discrimination. In particular, we assess to what extent the regulatory frameworks address the challenges of algorithmic transparency and accountability. We argue that while the law regulates both algorithms and their discriminatory effects, the framework is insufficient in addressing the complex interactions that must take place between system developers, users, oversight and profiled individuals to fully guarantee algorithmic transparency and accountability.