The rapidly growing development of intelligent autonomous and automated weapon systems has become one of the most controversial sources of debate in the international arena. One of the many concerns that emerges about this technology is the existence of an accountability gap. This fear stems from the complexity of holding a human operator responsible for the possible malfunction of a weapon system; therefore, when these crimes are not intentional, the question arises as to who should be held criminally responsible for serious violations of international humanitarian law. Therefore, the aim of the present article is to examine the criminal liability arising from serious violations of international humanitarian law by intelligent autonomous weapon systems as a war crime, carried out qualitatively and based on inferences from sources and texts, and in a descriptive-analytical manner. The findings and conclusions indicate that lethal autonomous weapon systems may cause serious violations of international humanitarian law, mainly because their target selection is highly unpredictable. Under current international criminal law, criminal liability can only be attributed when there is at least some intent and knowledge of the crime. The mere possibility of a violation is not sufficient to hold operators or programmers responsible for war crimes committed by lethal autonomous weapon systems. Command responsibility seems to be the most appropriate mode of liability, provided that these weapons are deployed under a specialized command structure. The command structure should include a group with technical expertise to test and continuously monitor the operation of the weapon systems.