This project focuses on the predictive AI technologies used to identify the geographic areas where policing resources should be targeted and also on the first component of the criminal justice process which is the proactive or preventative policing phase when the police use data-driven predictive algorithms to identify the geographic areas to target for crime prevention purposes.
Although the project will focus on the proactive policing phase when decisions are made about how and where to target policing resources, data generated during the project can also be used for reactive policing activities such as criminal investigation and other criminal justice activities that are the focus of other ATI projects.
There is growing interest in exploring the benefits and potential harms of predictive algorithms in criminal justice systems, but the extant empirical literature focuses mainly on policing and sentencing contexts in the US and other jurisdictions. Limited independent research exists on the nature and impact of the predictive algorithms employed by police services in the UK. The proposed project seeks to address this gap in knowledge by studying the predictive mechanisms underpinning the PredPol Machine Learning (ML) algorithm that is used by some policing services in the US
and by Kent Police in the UK.
Principal Investigator: Dr Pamela Ugwudike (Southampton)