Artificial Intelligence and Issuing the Criminal Sentences: Decision-making or Decision-taking?

The role of artificial intelligence-related technologies in legal knowledge is significant like most other scientific fields. Legal criminal sciences, and criminal justice system, have not unbenefited from the presence of this technology. In many developed countries, artificial intelligence-related systems are applied in various stages of criminal proceedings. The basic question is, "are the technologies related to artificial intelligence applicable in the process of issuing a criminal sentence? And if yes, what are the challenges in front of it?" The method of research is descriptive and analytical. It indicates that, the influence of personal feelings in judicial proceedings, lack of accuracy and speed in handling cases, application of prejudices and individual tastes, as well as conscious or unconscious biases, are among the most important reasons of applying artificial intelligence technologies at sentencing level. However, using the cloak of judgment in an autonomous is the main barrier of this path and will bring harmful challenges, such as partiality, lack of transparency, dehumanization in the decision-making process, and homogenization in judgment. On this basis, the presence of artificial intelligence as a decision-making tool, facilitator and assistant in the criminal sentencing phase is recommended. However, the use of this technology in a decision-taking and independent manner is contrary to the modern requirements of judgment and punishment doctrine. In fact, it will bring legal breaks and question the human contribution in the administration of justice.