Criminal Liability for Homicide Caused by Artificial Intelligence

Abstract

The expanding use of artificial intelligence in autonomous transportation, healthcare, industry, robotics, and decision-support systems has generated new questions concerning the attribution of criminal liability when harmful consequences, particularly human death, result from the operation of intelligent systems. The central problem arises when a fatal outcome is directly produced by an artificial intelligence system, while multiple natural and legal persons are involved in its design, production, programming, training, deployment, operation, and supervision. This study adopts a descriptive-analytical method to examine the foundations and criteria of criminal attribution in cases of homicide caused by artificial intelligence and analyzes the potential liability of programmers, designers, manufacturers, developing companies, owners, operators, and users. The findings indicate that the technical autonomy of an intelligent system, by itself, cannot justify recognizing artificial intelligence as an independent criminal offender. Under current legal conditions, criminal liability should instead be determined according to the role of human and corporate actors in creating, controlling, or failing to prevent the relevant risk. In this regard, factors such as foreseeability of harm, effective control over the system, access to technical information, the ability to prevent the fatal outcome, breach of a duty of care, and the degree of causal contribution of each actor are of fundamental importance. Where several factors jointly contribute to death, responsibility may need to be distributed among multiple actors rather than attributed exclusively to one person. The study concludes that general principles of criminal law remain applicable to artificial-intelligence-related homicide, but they are not entirely sufficient for autonomous and high-risk systems. Specific legal rules concerning risk assessment, human oversight, algorithmic traceability, safety obligations, post-deployment monitoring, and corporate criminal liability are therefore necessary to prevent responsibility gaps and ensure effective legal accountability.