ARTIFICIAL INTELLIGENCE BETWEEN CRIMINAL ACCOUNTABILITY AND THE DIFFICULTY OF ATTRIBUTION (MEDICAL ROBOT AS A CASE STUDY)

AI entities, particularly medical robots, face a fundamental challenge in criminal accountability: the difficulty of attributing criminal error to a specific entity when harm occurs to a patient. When AI algorithms make an incorrect treatment decision that leads to health complications or death, it is difficult to easily determine whether the error was the result of a programming error, poor device maintenance, or even human misuse. The criminal liability of AI entities, particularly medical robots, hinges on the issue of whether these entities have legal personality. This remains controversial in jurisprudence, given the absence of legislation. In this context, the human agent theory has emerged, according to which criminal liability is distributed among the manufacturer, programmer, and user of AI entities.