Michael Law-Smith (University of Ottawa – Faculty of Law; University of Toronto – Department of Philosophy) has posted Algorithmic Sentencing and Human Dignity on SSRN. Here is the abstract:
This article argues that the use of algorithmic sentencing in the criminal justice system can be in tension with the state’s duty to respect the human dignity of offenders. Sentencing law in many legal systems requires judges to make complex normative judgments about a person, their conduct, and how the state may treat them in light of that conduct. These judgments implicate an offender’s dignity because they engage with their self-understanding as a moral agent and their sense of worth as a human being. However, algorithms and other forms of predictive artificial intelligence do not engage in anything like the kind of normative reasoning that sentencing law demands. Accordingly, when the state substitutes one or more elements of this normative reasoning with an algorithm’s output, it risks failing to respect the offender’s moral status. Treating the normative judgments sentencing law demands as mere empirical predictions is not only a category mistake but a failure to justify the sentence and to recognize the offender as a moral agent.
Recommended.
To receive a daily summary of posts from Legal Theory Blog by email, get a free subscription to Legal Theory Stack.