The purpose of this paper is to provide a theoretical foundation for the possibility of ethical symbiosis with human and constructive artificial Intelligence, based on the autopoiesis theory of Humberto R. Maturana, Francisco J. Varela and Niklas Luhmann. This paper will clarify a structural coupling with a social environment by examining the social participation and behavior of artificial intelligence, a symbol of the development of science and technology in modern society. According to the autopoiesis theory, there is not already an objective world to be recognized, but the world and recognition actors are formed together in the process of cognition. And the self-generating system is recursive and reflective, creating norms and rules for members to act together and maintaining safety. In the functional aspect of the self-production system (quasi) autonomous artificial intelligence and autonomous human beings can co-evolve through structural coupling and create a world together. The behavior of artificial intelligence is also that it should be reproduced so that action adjustments can occur in the interrelationship of what happens to the receiver. According to the autopoiesis theory, artificial intelligence s behavior should also be based on reflective cognition, which takes place within language, as should the responsibility of conduct be based on the equality between cognition and action.