Innocent Until Predicted Guilty? Artificial Intelligence and Police Decision-Making
Alexander Babuta
29 March 201814 Minute Read
Durham Constabulary’s Harm Assessment Risk Tool (HART) uses random forest forecasting to predict an individual’s risk of re-offending based on 34 variables. Based on this prediction, individuals are then classified into high, moderate or low risk groups.
The use of artificial intelligence and algorithms has the potential to enhance various aspects of criminal justice decision-making. However, these capabilities bring with them a number of complex ethical and legal considerations, which are at present quite poorly understood.
Please wait while we verify your account can view this article.
