Problems of introducing information systems based on artificial intelligence into the activities of law enforcement agencies

The article is devoted to the analysis of problems related to the introduction of information systems based on artificial intelligence into the activities of law enforcement agencies. The positive aspects of using artificial intelligence systems to solve official tasks by law enforcement officers are emphasized. The directions for the application of AI systems in law enforcement activities are described. At the same time, a range of current and potential problems related to the use of intelligent systems that already serve the law enforcement sphere and may be used in the future are outlined. It is emphasized that artificial intelligence systems developed for law enforcement officers have a high degree of risk, threatening the protection of human rights. It is assumed that the actions of law enforcement officers as a result of the use of AI systems may not comply with the principle of proportionality. The emphasis is on the issues of the quality of AI developments and the data on which they are trained, and the related problems of liability and transparency policy of the manufacturer, as well as the problems of intellectual property protection. It is emphasized that the use of inappropriate AI systems in procedural actions is unacceptable, as it may violate the human right to protection and a fair trial. The opinion is affirmed that artificial intelligence systems designed for law enforcement agencies must be accurate and reliable; their intellectual work must be transparent, understandable, and safe. Attention is drawn to the need to develop high-quality AI systems with narrow specialization, designed to help law enforcement agencies perform certain tasks: risk assessment, detection of disinformation, detection of deepfakes, profile creation, for analytics of crimes against individuals, etc. It is proposed to classify not only AI systems for law enforcement agencies, justice, and border services, but also those developed for customs and tax authorities as high-risk. The problem of using intelligent systems to combat crime in the metaverse is outlined. The problem is identified as the absence of regulatory acts in national legislation regulating the implementation and use of information systems based on artificial intelligence in law enforcement and other areas of life.