The article maps the ethics of AI in criminal sentencing. Several jurisdictions have already begun to use AI to assist human judges in sentencing, and more jurisdictions are expected to follow in coming years. With the development ongoing, there is an urgent need for greater clarity on the ethical dimensions of AI sentencing. This article seeks to take stock of the scholarly debate that has taken place so-far. Based on the many contributions to the debate, the article identifies 13 ethical problems that AI sentencing raises. The article contains four maps that show different ways of grouping the 13 ethical problems. Some of the ethical problems relate to the outcome of sentencing, some to the process, and some to the overall system of criminal justice. Some ethical problems arise only for autonomous AI and not for advisory AI. And some ethical problems require solutions based in penal ethics, while others are general ethical problems shared by other uses of AI in public governance.
David Chelsom Vogt (Mon,) studied this question.