Can AI become Reliable Source to Support Human Decision Making in a Court Scene?

摘要

Recently, advances in artificial intelligence (AI) have provided support to human decision-making, and there has been some controversy regarding whether AI can be used to support judges in court. This study investigates the influence of data that is presented by a robotic system on human judgment in a sentence estimation situation in court. The study includes an experiment in which participants played the role of a juror in a sentence estimation task. Participants were presented with scripts of the case, reference materials regarding similar cases, and sentences given by the expert robotic system and human, and then, determined an adequate sentence based on this information. Results show that participants interacting with the robotic system tend to agree with the presented material (in the same way as with human experts) when it is adequate. This result shows that robotic systems are treated the same way as humans, and that intelligent systems can be a reliable source of information for human decision-making in court. Moreover, the results show that robotic systems may, in some cases, be considered more trustworthy than humans.