With the expansion of artificial intelligence and its increasing autonomy, new legal challenges have been raised in the field of criminal liability resulting from the harmful actions of these systems. In the event of a crime being committed by autonomous artificial intelligence, determining responsibility and attributing it to a specific person or entity is faced with serious ambiguities. Therefore, research in this field is necessary to examine the jurisprudential and legal foundations to provide appropriate solutions for determining criminal liability and compensating for damages resulting from the actions of artificial intelligence and to protect the rights and interests of society against the possible dangers of this new technology. The present study, using a descriptive-analytical method, has reached the conclusion that in limited artificial intelligence, responsibility lies with human factors, and at more advanced levels, this becomes more complicated. From a jurisprudential and legal perspective, rules such as harmlessness, expediency, and justice require the need to compensate for damage and punish the harm caused. Proving the elements of a crime and the fault of human factors related to artificial intelligence is a way to attribute responsibility. Finally, the formulation of new laws, while supporting innovation, must also guarantee the rights of society.