Sentencing as a human process? Human judges’ experiences of machine-based sentencing

Abstract Machine decision-making technologies, such as artificial intelligence, are increasingly gaining prominence in sentencing. While criminologists have expressed concerns about these technologies, the ways in which sentencing judges perceive such developments remain underexplored. Drawing on interviews with 24 Israeli district court judges, this study identifies two dominant themes shaping judicial attitudes toward machine-based sentencing (MBS). The first is held by techno-optimists who view sentencing as a normative reasoning process that could be amenable to automation, while the second is advocated by techno-pessimists who conceive sentencing as a phenomenological process that requires interpersonal-emotional engagement as well as moral deliberation, and therefore cannot be reduced to algorithmic analysis. The judicial imagination of MBS reveals the ways in which many judges resist the managerial, rational and bureaucratic logic that often underpins criminological visions of “robots judges.” They illuminate that the feasibility of MBS is deeply rooted in the ethical meaning of sentencing and judicial professional identity and legitimacy.