This report describes the case of Criminal Sentencing AI technologies as a research project and outlines the background and the rationale to suggest a research plan. There is a rise in criminal sentencing AI technologies within the court system. There are eight states with documented uses, by including the recidivism score of the individual within the sentencing process. Those states are Arizona, Colorado, Delaware, Florida, Kentucky, Louisiana, Oklahoma, Virginia, Washington, and Wisconsin. While a promising technology AI can lead to inheriting biases of those who create them. When deciding the sentencing of an individual, it’s important to think about the factual data and also the context of the situations that occurred with that particular incident. By employing and relying specifically on criminal sentencing technologies, we eliminate due process, while also removing the “human” from the judicial system, where it’s already minimal. The question becomes, are these tools harming individuals negatively at a disproportionate rate versus others? Additionally, whether these tools are actually helpful or successful in predicting the possibility of recidivism? In this research project, I plan to evaluate the effectiveness of the tool Compas, while also looking at whether there is a need for “the human” in the judicial system, particularly, with the sentencing of individuals.
Back to Top