The Risks And Regulation Of Predictive Policing

Predictive policing has been used as a “governance tool” in China for a decade,but as a“governance object,” it has only recently entered the research field.There are three major risks associated with predictive policing: prediction accuracy,service normativity,and social adaptability.The risk of prediction accuracy falls into three categories: speculative risk,acceptable risk,and delay risk,with characteristics that include difficulty in calculation,dynamic evolution,and easy spread.The lack of basic data,defects in predictive models,and the highly complex nature of human activities can cause risks to prediction accuracy,leading to false positive and false negative results,biases,and other accidents,which can result in ineffective prediction,incorrect crackdowns,and misallocation of resources.The risk of service normativity falls into three categories: process risk,independent risk,and critical risk,with relatively hidden,continuously growing and non-dispersible characteristics.Poor selection of predictive conclusions,chaotic execution of predictive conclusions,and blind trust in predictive policing by police officers can lead to risks of service normativity,causing accidents such as excessive policing,excessive law enforcement,discriminatory law enforcement,and gray law enforcement,disrupting daily order,damaging police-civilian relations,and generating legal liabilities.The risk of social adaptability falls into three categories: strategic risk,transformation risk,and political risk,with negative externalities,cross-coupling,and amplification characteristics.The uncertain return on investment,legitimacy,and the creation of social anxiety in predictive policing can lead to risks of social adaptability,causing accidents such as public resistance to predictive policing,legislative prohibition of predictive policing,police suspension of predictive software,and verdicts of illegal predictive policing,resulting in reductions in the scale of predictive policing,technological downgrades,and stigmatization.To regulate the risks associated with predictive policing,three basic concepts are suitable: scientific rationalism,relative rationalism,and prudent optimism.To regulate the actual risks of predictive policing,three paths should be followed: technology,norms,and ethics.Regarding technology,it is necessary to implement full lifecycle management of basic data,invite academic and corporate partners to jointly develop systems with law enforcement agencies,promote the simplification and visualization of predictive algorithm models,introduce third-party review and evaluation of predictive systems,and provide adaptive training to police officers on the use of predictive systems.Regarding norms,it is advisable to establish specialized authorization clauses for predictive policing,establish the presumption of harmlessness for preventive interventions,establish a manual screening and evaluation system for predictive conclusions,improve the standardized management and disposal of predictive conclusions,and improve the system of police information disclosure and modification.Regarding ethics,it is advisable to lead the operation of predictive policing with caring ethics,guide the design of algorithm models with moral principles,promote the governance of public safety risks with flexible means,and promote social consensus through democratic consultation.