Features of the use of artificial intelligence in forensic activities: some relevant questions

The article examines the role of neural networks, including recurrent and convolutional neural networks, in ensuring efficient data processing and decision-making in forensic tasks. The findings highlight the significant potential of artificial intelligence for optimizing forensic research. The development and implementation of effective technological solutions combined with legal mechanisms can optimize the detection and investigation of crimes, as well as the conduct of expert research. In the context of the rapid development of technological progress, innovative technologies permeate all spheres of human activity. Forensic science, being an integral part of the law enforcement system, does not remain aloof from these trends. Law enforcement agencies face a critically important task: to research and implement the latest technological solutions that contribute to improving the effectiveness of forensic activities. Such solutions open up a wide range of opportunities for improving crime detection and investigation methods, improving the quality of expert research, and improving the justice system as a whole. A neural network is a computing device consisting of interconnected nodes organized into multi-layered structures capable of learning from large amounts of data. Optimization of the learning process is achieved through the use of complex algorithms that ensure the effective operation of this network in various tasks. Neural networks are classified according to different architectures, but recurrent neural networks (hereinafter referred to as RNNs) have demonstrated optimal efficiency in tasks related to optimizing the detection and investigation of crimes. RNNs are a type of neural networks that are used in various tasks, including speech and voice recognition, time series prediction, and natural language processing. RNNs were created to overcome the limitations of classical feed-forward neural networks, which could not efficiently process sequential data. Unlike traditional neural networks, RNNs have recurrent connections that allow them to store information about previous inputs and take it into account when processing the current input. RNNs are distinguished by their ability to process data sequentially, storing information about previous elements in their internal memory. This property makes RNNs suitable for processing tasks involving time dependence, such as natural language processing and speech recognition. In conclusion, it should be concluded that it is necessary to take into account the fact of continuous improvement and methods of committing crimes, including through the use of artificial intelligence systems by criminals, as well as making mistakes by the platform (identifying a legitimate transaction as suspicious, which, in turn, may lead to a violation of the rights of consumers of banking services). Therefore, it is necessary to continuously optimize the platform for monitoring transfers of Russian citizens to identify suspicious transactions. The issue of legal protection of personal data also requires a legal rethink. The authors declare no conflicts of interests.

Keywords

Authors

References