Introduction. Digitalization fully represents itself not only in business or state governance, but also in the field of jurisprudence. Digitalization of the latter can be discussed in three aspects: digitalization of law practices, science and education. At present, a number of routine legal operations are being digitalized – civil contract drafting constructors are being created, legal data reference systems are being developed. Digitalization opens up the great prospects for legal education, for example, in extended use of computer games in teaching students. The least studied are the trends and prospects of digitalization in the Russian juridical science, which are represented in the use of text and graphic neural networks (for example, ChatGPT). However, the use of the neural networks has a number of ethical and legal restraints that are little discussed in the present-day juridical science. Therefore, the aim of this article is to reveal the trends and prospects of using the neural networks in scientific juridical research, as well as to identify the advantages and threats emanating from the modern neural networks (on the example of ChatGPT).

Materials and Methods. The research was carried out using the methods of scientific cognition commonly accepted in Russia: general scientific (dialectical) method and specific scientific methods (analysis, synthesis, concrete-historical, logical, legal cybernetics method, systemic methods and others). The neural network research method, which enabled to analyse the algorithms of the ethically acceptable usage of ChatGPT neural network, was also used in the article. According to the methodology chosen by the authors, the institutions, programs and technologies were not investigated.

Results. Within the research, the authors have analysed the theory and practice of using the neural networks for conducting the scientific research in the environmental law, the results of using the graphic neural networks have been shown on the certain examples (two figures generated by the neural network are attached). In the framework of using the text neural networks, the algorithms for their application have been proposed (on the example of conducting a comparative analysis of a number of legal concepts, including the sustainable development and green economy). The gaps in the legal regulation of the use of ChatGPT neural network have been revealed and ethical parameters for its use have been determined. Using the neural networks in the context of the civil law regime has been investigated, the recommendations have been formulated as a response to using the neural networks in drafting the contracts and in some other fields of juridically significant activity.

Discussion and Conclusion. Nowadays, digitalization affects all spheres of public life, including the juridical science and legal education. Of all the existing digital technologies, we have chosen the artificial intelligence technology (ChatGPT) as the most well-known and, at the same time, having the great prospects. This technology can be useful both in the field of legal education and juridical science. ChatGPT (graphic) can facilitate the process of teaching the students at law faculties, by improving the methodological support of lectures and seminars (presentations, graphics). In terms of juridical science, ChatGPT (text) can be useful for summarizing the scientific and/or empirical materials, conducting classifications, adjusting the conclusions, etc. It is worth emphasizing that the direct use of the texts generated by the artificial intelligence is unethical, and is directly prohibited in a number of scientific journals. However, a complete ban on this technology is not expedient. It would be more forward-looking to define the exact parameters of the neural network use in the juridical science and legal education.