"I robot”: the criminal liability of Artificial Intelligences

Humanity is on the threshold of an era in which robot, androids and other manifestations of artificial intelligence seem to be about to launch a new industrial revolution, susceptible to affect all social strata and many sectors of the human life, making essential that the legislation considers all implications of it. The technological evolution has led to the construction of artificial intelligences with increasing autonomy, capable of making decisions, interacting with the environment and activating themselves, independently of human intervention. Such technological development raises the question of the possibility to recognize a legal personality to artificial intelligences, considering them legal entities with peculiar characteristics. The answer to this question has significant implications in the criminal field, where there is a need to establish whether a personal criminal liability of the robot can be configured. This would require a reconsideration of all the classic categories of criminal law such as conduct, causality, author and mens rea. Otherwise, if it is considered that robots are not currently comparable to legal entities, due to the lack of actual free will, the criminal liability should be attributed to their creators. In this case, excluding the hypotheses of a voluntary use of the robot for the commission of crimes, it is necessary to make a reflection that involves various paradigms of criminal law from causality to culpability. In this context emerges the need of a legislative intervention that adapts the traditional categories of criminal law to the changing technological context, keeping in mind the respect of fundamental principles of criminal law as the rehabilitative function of punishment and the principle of extrema ratio.