Ethics Documents in the Field of AI. Concepts, Achievements and Problems

The development of AI systems and their widespread use in various spheres of people’s lives and society have given rise to moral problems. As a result, a new field of applied ethics emerges - AI Ethics, which consists not only of scientific research into the ethical aspects of the design and use of AI systems but has practical implementation in various ethical documents: declarations, recommendations, ethical codes and ethical standards. In this article, in the course of studying existing ethical documents, a typology has been developed based on differences in their statuses, sources of origin and moral agency. The analysis made it possible to identify the main ethical issues discussed in AI technologies and identify the most frequently used concepts and terms. Particular attention is paid to the transparency, privacy, trust. The article formulates significant points of discrepancy in the understanding of these terms in various AI ethics documents. As a main conclusion, we can formulate the idea that currently most ethical documents have completed the stage associated with identifying the main problems and concepts of AI ethics, although there is some disagreement about their understanding. It is noted that the vast majority of ethical documents analyzed do not sufficiently develop the practical component, which should be the basis for making decisions to prevent and resolve moral problems arising in connection with the design and use of AI.