Feasibility of AI-powered war weapons criminal liability and the challenge of impunity in the International Criminal Court

Abstract

he issue of delegating the authority to kill humans to artificial intelligence in wartime is one of the most challenging emerging disciplines, as it raises numerous legal issues. This is the first time that legal knowledge has encountered a phenomenon that, despite being created by humans, is independent of humans in thinking and decision-making, and it may be possible after multiple stages that even the creator of the system cannot comprehend or control its operation. Therefore, this article discusses the obstacles that the International Criminal Court must overcome to apply direct criminal responsibility and the doctrine of commander's responsibility in cases where artificial intelligence weapons have led to the killing of civilians and war crimes, as well as the steps that should be taken to overcome these obstacles. Following a documentary approach and a descriptive-analytical analysis, this article concludes that the current provisions of the statute regarding artificial intelligence need to be revised due to the emergence of a concept called the gap in responsibility, which is caused by the unique nature of artificial intelligence. To adapt to the new conditions, the international community must require manufacturers to provide maximum transparency and prohibit producing certain types of artificial intelligence. Regarding the possible killings caused by the use of artificial intelligence weapons with the ability to self-learning after production, the court statute must explicitly and exclusively specify the rules of criminal responsibility so that the existence of AI weapons does not lead to impunity.

Keywords

• artificial intelligence
• command responsibility
• autonomous weapons
• impunity