Feasibility study of civil liability of artificial general intelligence due to damage in civil law

Advanced artificial intelligence, based on its knowledge and experience, can intend and perform various actions to achieve its goal. In this article, first, the identity and definition of the technology that creates artificial intelligence is discussed. Then the examples of losses due to its independent act are stated and analyzed. In case of harm to a person due to those actions, according to the theory, artificial intelligence can be considered responsible. Because it is conscious and also does not represent another person and has harmed independence. In the following, it is determined that the creators, users and owners or developers of artificial intelligence are also responsible and their level of responsibility is determined. According to the Iranian legal system, artificial intelligence cannot be called a person because it is not subject to the current definition of a natural person or a legal person. To establish the basis in artificial intelligence liability, civil liability caused by damages to animals and property of owners, minor and insane guardians, and agency theory are analyzed.