A study on the analysis of some of the problems that arise when discussing the possibility of establishing codes of conduct in connection with the development of autonomous intelligent systems. First, it is a problem of the normative status of the requirements of the ethical code, which becomes part of the software algorithm. Secondly, it is a question of the subjectivity of the code of ethics. Thirdly, this is a problem that is related to the need to develop special norms for the «morality of machines».