Search for a command to run...
This article provides a comprehensive analysis of the application and legal regulation of artificial intelligence systems in US criminal proceedings. The relevance of this research is driven by ongoing digital transformation, in which the introduction of artificial intelligence systems into criminal proceedings affects fundamental rights and freedoms, changing traditional decision-making procedures. The significance of this work is determined by the need for the Russian Federation to understand the American experience, which represents a large-scale natural experiment in the integration of technologies within a decentralized legal system and the dominance of the private sector in the development of specialized artificial intelligence systems. The purpose of this study is to analyze law enforcement practice and the emerging regulatory ecosystem, as well as to formulate conclusions of theoretical and practical value for the development of the domestic legal system. The study utilizes a combination of general and specialized scientific methods, including the analysis of regulatory acts, case law, and scientific literature, comparative legal analysis, as well as a method for systematizing and classifying empirical data related to the implementation of artificial intelligence systems in the United States. The scientific novelty of this study lies in its systematic analysis of the American model for regulating the use of artificial intelligence in criminal proceedings as a uniquely "reactive" and decentralized system, shaped primarily by judicial precedents and state legislation. The article critically examines key issues: algorithmic and ideological bias; the explainability of decisions made by artificial intelligence systems; the opacity of algorithms; and the risk of "blind delegation" of decisions to the system. The main conclusions of the work boil down to the need for preemptive federal legal regulation in Russia, which should establish a classification of AI systems by risk level, mandatory audit requirements, procedural guarantees of the rights of the parties, and the allocation of liability. The author substantiates the priority development of domestic AI systems, the phased implementation of AI, beginning with a "co-pilot" model, where AI systems perform an auxiliary analytical function, as well as the need to develop a professional culture and specialized training for lawyers to critically analyze the results generated by artificial intelligence systems.