Judges Using Algorithms to Assist in Decision Making Process
A recent experiment conducted in US courtrooms has revealed that a popular risk assessment algorithm may provide harsher recommendations compared to human judges. This discrepancy could be attributed to the algorithm’s limitations in accurately predicting which defendants are likely to violate pretrial agreements.
Harvard University’s Kosuke Imai explains, “Some jurisdictions sought our assistance to evaluate the effectiveness of these algorithmic recommendations in aiding judges to make informed decisions.”
Within the US criminal justice system, judges are tasked with determining whether defendants should remain at home or in…