The advanced programming being used in some machines that operate on Artificial Intelligence (AI) has demonstrated capabilities that are dangerous enough to build self-experience that enables them to make unilateral decisions, leading to them getting out of control. Some incidents were reported that led to the killing of people due to AI systems. This has raised public concerns about these machines that may commit crimes and the extent to which they are legally accountable for their actions. Therefore, this study aims at studying the extent to which AI machines could commit criminal actions, especially machines that have the ability to make their decisions independently such as robots, and to determine criminal accountability for crimes committed by AI systems, by determining the personality of the AI entity being accountable for the criminal acts committed by AI systems? The researcher has followed the analytical and comparative approaches for analyzing the study problem, highlighting the legal aspects related to AI crimes and their criminal accountability, to reach the best solutions that can be used and applied legally. This study has reached a set of results, the most important of which is that AI systems have demonstrated advanced programming in some machines that use AI, which amounts to the risk of building self-experience that enables them to make ي unilateral decisions, leading to them getting out of control. This has raised public concerns about these systems committing crimes without being legally held accountable for their actions. In addition, AI systems do not have criminal capacity or legal personality, in order to be criminally accountable. This does not negate the criminal accountability of the human personality standing behind the AI system. AI technologies will continue to develop in new forms, and some of them may cause harm and criminal behavior. This requires criminal legislators to respond quickly to develop legal and regulatory frameworks to confront these concerns. This study has also come out with several recommendations, mainly, the need to establish legal frameworks to determine the AI criminal liability and the need to address potential criminal acts of AI entities, as well as imposing legal restrictions on programmers and users of AI systems.