Potentials and Possible Drawbacks of Adopting Predictive Algorithm AI in the Ethiopian Criminal Justice System: Lessons from the US Experience

Abstract

Predictive algorithm AI is a type of machine learning that predicts future events by using data and some variables. Predictive algorithms are being applied in different sectors including the criminal justice system such as in predicting crime before it happens, re-offending, flight risk, or recidivism. Ethiopia would benefit from adopting predictive algorithms in the criminal justice system by carefully analyzing the potential and possible drawbacks. The objective of this research is to scrutinize the potential and possible drawbacks of adopting the algorithms in the Ethiopian criminal justice system taking the United States’(US) experience as a lesson. The study employs a qualitative research method with a comparative analysis taking the US as a case study. The US is opted for because the country has a long history of implementing predictive algorithms with a record of the evaluation of the application of the systems in various domains. Examination of the US experience shows that predictive algorithms have immense benefits and some drawbacks mainly related to the data set and the design of the models and best experiences by minimizing the drawbacks and maximizing the benefits can be taken to Ethiopia. The writer suggests Ethiopia's historical, political, and cultural context have to be considered while examining the legal, ethical, and social ramifications of using predictive algorithms. This study recommends the adoption of predictive algorithms with the right design, implementation, and evaluation in place, the adoption of legal frameworks that govern the usage of the systems, and comprehensive data protection law, and the establishment of proper infrastructure.