Legal aspects of artificial intelligence in ensuring national security: challenges and threats

The subject. Artificial intelligence (AI) opens up wide opportunities for strengthening national security, but also raises a number of legal and ethical issues that need to be carefully analyzed and resolved. The implementation of a government strategy aimed at improving tools and approaches to information protection using AI is a key aspect of ensuring national security and maintaining law and order.The aim of the article was to confirm the hypothesis that ensuring national security depends on algorithms implemented in artificial intelligence (AI) technologies.Methodology. The criteria formulated in the President Decree of the Russian Federation No. 124 are fundamental on the basis of which trusted AI systems are developed. The analysis of legal acts, acts of technical regulation and scientific literature was made.Main results. In areas where there is a threat to national security, the use of trusted AI technologies is becoming mandatory. In accordance with Presidential Decree No. 124, trusted AI technologies are defined as meeting safety standards and developed based on the principles of objectivity, non-discrimination and ethics. Their use should exclude the possibility of harming a person, violating his rights and freedoms, as well as harming the interests of society and the state. Only two principles formulated in that Decree are disclosed – this is the principle of non-discrimination and objectivity. Despite the use of the term "ethics of norms", "ethics" in the legislation in the field of AI, ethics is not defined as an independent principle. Ethics in the field of AI can be understood as a system of views and principles aimed at protecting the moral and social consequences associated with the use of various data processing models contained in AI technologies. The principle of objectivity or impartiality is related to the ability of the system to inspire confidence and exclude unjustified bias of the formed estimates. The principle of objectivity is closely intertwined with the principles of transparency of AI systems and data protection and security. To achieve the objectivity of data processing by AI systems, it is necessary to use high-quality and representative datasets obtained from reliable sources. The implementation of the principle of non-discrimination contains certain contradictions that manifest themselves in the conditions of ensuring national security. Since, despite the fact that it is intended to exclude the possibility of using AI algorithms for processing and analyzing data containing discriminatory criteria, if necessary, the state can apply other criteria to ensure the national security.Conclusions. The implementation of the above principles, the creation of a trusted environment for the use of AI technologies, and the development of trusted AI technologies all ensure information security, the effectiveness of the system for protecting the interests of citizens, society, information, and information infrastructure, thereby ensuring national security.