Artificial Intelligence in Forensic Expertology
E. V. Chesnokova,A. I. Usov,G. G. Omel’yanyuk,M. V. Nikulina
DOI: https://doi.org/10.30764/1819-2785-2023-3-60-77
2023-11-09
Theory and Practice of Forensic Science
Abstract:The article reviews the issues of studying the capabilities and areas of application of artificial intelligence technologies (AI) in forensic expertology as a science of forensic examination and forensic expert activity. The authors have developed the criteria for choosing the direction for the introduction of AI in the system of forensic expertology, and specifically in its innovative section – the management system of forensic expert activity.The prospects for the development of AI technologies in forensic examination are associated with the processing of big data based on the expert situation, the reliability of the data used in AI training and further validation (assessment of suitability) of the applied training methodology, analysis of the results of AI technology. To solve organizational and legal issues of integrating AI technologies into legal proceedings and, specifically, into forensic examination, a system of standards regulating the order, algorithms and procedures for its implementation and use of is proposed. At the same time, the assessment of the suitability of the results of the use of AI in forensic examination should become an ongoing process included in the activities of each forensic expert organization. The necessary consistency of this process determines the updated paradigm of forensic examination in the conditions of functioning AI technologies in it and the cyclical nature of the entire process of their implementation and use. Process cycle is a set of sequential actions at different levels: initiation of AI technology, evaluation (suitability) of its results at the first level, adjustment and implementation of the updated version of AI technology, assessment of the next level, etc. In GOST accredited ISO/IEC 17025-2019 forensic laboratories such system activity, although not related to AI, has already been implemented, it corresponds to the PDCA cycle (P – plan, D – do, C – check study, A – act). Therefore, for such laboratories, the modification of the organization and management of the activities in the implementation of AI technologies is the most organic and acceptable. In addition, all work on AI in the field of forensic expertise should take into account the provisions of the AI Code of Ethics, which applies to relations associated with the ethical aspects of creation (design, construction, piloting) the introduction and use of AI technologies at all stages of forensic activities.