Comment by Fair Trials

Today, Fair Trials has launched an interactive online tool designed to show how the police and other criminal justice authorities are using predictive systems to profile people and areas as criminal, even before alleged crimes have occurred. Our research has shown that more and more police forces and criminal justice authorities across Europe are using automated and data-driven systems, including artificial intelligence (AI), to profile people and try and ‘predict their ‘risk’ of committing a crime in the future, as well as profile areas to ‘predict’ whether crime will occur there in future. These systems are created and operated using discriminatory data, reinforcing structural and institutional inequalities across society and leading to unjust and racist outcomes. These predictions can lead to surveillance, questioning and home raids, and can influence pre-trial detention and sentencing decisions. There is growing opposition to predictive policing and justice systems across Europe, with many organisations and some Members of the European Parliament (MEPs) supporting a ban. Fair Trials is calling on MEPs to ban predictive systems when they vote on the Artificial Intelligence Act in the upcoming months. We hope that our example predictive policing and justice tool will raise awareness of the discriminatory outcomes generated by these systems. Each question in our predictive tool directly relates to real data used by police and criminal justice authorities across Europe to profile people and places. Using information about someone’s school attendance, family circumstances, ethnicity and finances to decide if they could be a criminal is fundamentally discriminatory. The only way to protect people and their rights across Europe is to ban these systems.