Legal regulation of intellectual property protection for artificial neural networks
M. A. Ogarok · Proceedings of the Southwest State University Series History and Law · 2024
Relevance. Fundamental differences in the processes of development and maintenance of artificial neural networks from the algorithmic approach determine the relevance of the development of the existing scientific and methodological platform for solving technological and regulatory issues in the field of software development and legal protection of intellectual property of artificial intelligence systems. The purpose of the study is to formulate scientifically based conclusions that determine the directions for solving the scientific problem of legal protection of intellectual property on artificial neural networks. Objectives: to formulate a list of scientific tasks necessary for solving the scientific problem of legal protection of intellectual property rights to artificial neural networks; to develop proposals for improving the regulatory framework for registering intellectual property rights to artificial neural networks. Methodology. The methodological basis of scientific research was the dialectical method of understanding the phenomena and processes of the surrounding reality. Results. The obtained results of the study provide the opportunity to improve the regulatory framework establishing the basic principles of legal relations arising during the registration of intellectual property rights to artificial neural networks. Conclusion. The solution to the problem of legal protection of intellectual property on artificial neural networks should be based on the improvement of the existing regulatory framework for the registration of intellectual property rights, as well as regulatory acts that determine the rules for registering and securing intellectual property rights to arrays of training sample data and (or) arrays of trained models of artificial neural networks, as well as to the topology of integrated circuits of neuromorphic processors.