I think this Detain/Release simulation was a meaningful activity. The lack of information on the crimes and defendants made it difficult for me to make a fair decision. Who was the defendant? What was his/her family background? Was this his/her first offense? The reasons that the prosecutor and defendants gave were the same in several cases. For example, several defendants repeated, “I’m very ill. If I’m jailed, my illness might get worse” or “The state’s case is weak, and they know it. I want to fight these charges…”. In the process of making my decision, I started to put less weight on the reasons that both parties gave and put more focus on the defendant’s level of violence, likelihood to miss the trial and commit a crime again. This was an easier marker to determine whether the defendant should be released or detained. I also detained the defendants that were marked with high levels of violence and committing a crime. This made me think about the role that risk assessment has on aiding the judge’s decision making process. For me, I did not question the accuracy of the risk assessment because it was the only reliable marker that I had to make a decision on. I had to make a decision for each case and the risk assessment was my “counsel”. 

Upon reflection, this makes me question the level of power we have given to AI and algorithms. Most of us fail to question the accuracy and authenticity of the software used in the government and legal institutions. We transfer power to these algorithms and software and allow these tools to affect our biases and decision making. I think that these algorithmic risk assessments offer institutions a standardized and “fair” counsel that is “objective”, time efficient and convenient. In this case, these software allows the judge and the judicial system to process more cases in a time-efficient and “objective” manner.

Porcaro, K. (2019, January 8th). Detain/Release: simulating algorithmic risk assessments at pretrial. Berkman Klein Centre. https://medium.com/berkman-klein-center/detain-release-simulating-algorithmic-risk-assessments-at-pretrial-375270657819