Legal Issues of Copyright Objects Processing by AI Systems in the Process of Machine Learning

Abstract

The research paper analyses the legal aspects of copyright objects processing by AI systems, including the legality of automated analysis of text and data in digital form — TDM (text and data mining) and machine learning. The research paper examines: the grounds for qualifying such processing as copyright infringement and certain obstacles to the full protection of copyright (including Big Data, being a subject of processing by AI systems). The paper proposes the adoption of certain public law principles of copyright objects processing by artificial intelligent systems, in particular: the principle of limited purpose, according to which the processing of works should be carried out exclusively for the purposes established by the operator of the artificial intellectual system; the principle of limited storage, which involves storing personal data (in a form accessible to identify the subjects of this data) no longer than required in-order to achieve the stated processing purposes; the principle of transparent reporting — reporting on the quantity and quality of processed data sets, accessible to any person, regardless of interest.