The Evolution of Open Licensing in the CIS and the Application of Artificial Intelligence Technologies to Processing Scientific and Technical Information

Introduction. Open licenses have become the foundation of the Open Access movement and open science. However, classic licenses, including the GNU GPL and the Creative Commons (CC) family, do not address the specifics of using works as training data for AI, nor do they regulate the legal status of neural network weights and the output data generated by models.

Aim. To analyze the evolution of open licensing in the CIS member states, identify the legal and practical aspects of applying artificial intelligence (AI) technologies to process scientific and technical information (STI) on leading regional platforms, and develop recommendations for adapting licensing mechanisms to the tasks of automatic abstracting, annotation, and thematic structuring.

Materials and Methods. The study is based on a comparative legal analysis of the national legislation of the Russian Federation, the Republic of Belarus, the Republic of Kazakhstan, and Ukraine regarding the regulation of open licenses. A review of the practical implementation of AI technologies (neurosearch, automatic classification, text summarization) on the platforms of the Scientific Electronic Library eLibrary, the National Electronic Library of the National Library of the Russian Academy of Sciences, the All-Russian Scientific and Technical Library of the Russian Academy of Sciences, the State Public Scientific and Technical Library of Russia, and the United Institute of Informatics and Problems of the National Academy of Sciences of Belarus was conducted. An analysis of the compatibility of Creative Commons (CC) licenses and specialized AI licenses (OpenMDW, RAIL) with machine learning tasks on NTI datasets was conducted.

Results. Heterogeneity of the legal regulation of open licenses in the CIS was established: the most developed systems are in Russia (Articles 1286.1 and 1368 of the Civil Code of the Russian Federation), Belarus (Article 45 of Law No. 262-Z), and Kazakhstan (Digital Code of 2026). It was found that platforms are actively implementing AI, but licensing restrictions create legal barriers to automated processing. Unresolved issues include the legal status of datasets, neural network model weights, and model outputs, as well as the lack of exceptions for text and data mining (TDM) in CIS legislation. It was shown that CC BY-licensed collections and public domain works are most suitable for creating open datasets.

Conclusion. To legally use AI in processing NTI, libraries and information centers are recommended to primarily use CC BY-licensed documents, comply with ND and NC restrictions, and label automatically generated content. Harmonization of CIS legislation, adaptation of specialized AI licenses to regional practices, and the development of a standard open licensing policy for automated processing of NTI at the level of the State Public Scientific and Technical Library of Russia are advisable.

1. Bornmann L., Haunschild R., Mutz R. Growth rates of modern science: a latent piecewise growth curve approach to model publication numbers from established and new literature databases. Humanities and Social Sciences Communications. 2021;8(1):(224). https://doi.org/10.1057/s41599-021-00903-w

2. Kumar R., Reddy M., Pappula P. Text Classification Performance Analysis on Machine Learning. MATTER: International Journal of Science and Technology. 2019;28(20):691–697.

3. Vakhrushev M.V., Goncharov M.V., Zasursky I.I. et al. Open access: History, current status and way to open science. St. Petersburg : Lan, 2020. 168 p. ISBN 978-5-8114-5034-3.

4. Yanochkina Y.V. Open access and open licensing by Russian journals. Scientific and Technical Libraries. 2024;(3):142-157. (In Russ.) https://doi.org/10.33186/1027-3689-2024-3-142-157

5. Fox J. et al. Open science, closed doors? Countering marginalization through an agenda for ethical, inclusive research in communication. Journal of Communication. 2021;71(5):764–784. https://doi.org/10.1093/JOC/JQAB029

6. Chuprakova P.A. The Institute of Open (Public) Licenses in National Copyright and Related Rights Legislation. Law in Modern Belarusian Society. 2023;(18):545–552. EDN: RDLJPE7

7. Romanenkova E. I., Sergo A. G. Legal Regulation оf Relations Concerning Free Licenses in Russian Law. Copyright (Bulletin of the Academy of Intellectual Property). 2020;(2):1–20. (In Russ.).

8. Stratan A., Bordian E. National infrastructure for technical and scientific information — a cornerstone of Open Science and research visibility in the Republic of Moldova. Information and Innovations. 2025;20(3):5–12. https://doi.org/10.31432/1994-2443.2025.14. EDN: DFFSAT

9. Stratan A. N., Guslikova N. I. The state of the scientific and technical information sphere in Moldova. International Information Forum. 2017;42(2):31–37. (In Russ.).

10. Lobodina Yu. A. Legal regulation of Creative Commons and open licenses: current status and development prospects. Law and Business. 2021;(3):48–55. (In Russ.). EDN: ZSRMZW

11. Moskalenko I.A. The Importance of Creative Commons Licenses for the Evolution of Copyright in Cross-Border Relations. Current Issues of Russian Law. 2021;16(9):141–148. (In Russ.). https://doi.org/10.17803/1994-1471.2021.130.9.141-148

12. Moiseeva N. A. Neural information retrieval in digital libraries to improve the quality of library and information services. Scientific and technical libraries. 2026;(4):66–87. (In Russ.). https://doi.org/10.33186/1027-3689-2026-4-66-87

13. Goncharov M. V., Sokolinsky K. E., Shraiberg Ya.L. Application of artificial intelligence in the practice of scientific and technical libraries: study of potential, experience of use, assessment of prospects. Scientific and technical libraries. 2025;(12):144–164. (In Russ.). https://doi.org/10.33186/1027-3689-2025-12-144-164

14. Melnikova E. V. Relevance of the application of artificial intelligence tools in modern scientometric research. Scientific and technical information. Series 1: Organization and methods of information work. 2024;(2):19–25. (In Russ.). https://doi.org/10.36535/0548-0019-2024-02-2

15. Kusakin I. K., Fedorets O. V., Romanov A. Y. Classification of Short Scientific Texts. Scientific and Technical Information Processing. 2023;50(3):176–183. https://doi.org/10.3103/S0147688223030024

16. Sohrab M. G., Asada M., Rikters M., Miwa M. BERT-NAR-BERT: A Non-Autoregressive Pre-Trained Sequence-to-Sequence Model Leveraging BERT Checkpoints. IEEE Access. 2024;12:23–33. https://doi.org/10.1109/access.2023.3346952.

17. Goncharov M. V., Kolosov K. A. Features of using large language models in compiling text abstracts. Scientific and technical libraries. 2025;(11):203–214. (In Russ.). https://doi.org/10.33186/1027-3689-2025-11-203-214

18. Bychkova E. F., Kolosov K. A. Analysis of the possibilities of automatic abstracting of articles using the example of sources from the database “Ecology: Science and Technology” of the State Public Scientific and Technical Library of Russia. Scientific and technical libraries. 2023;(10):99–120. (In Russ.). https://doi.org/10.33186/1027-3689-2023-10-99-120

19. Yadrevsky O. O. Main Directions for Improving Copyright Legislation in Light of the Development of Digital Technologies. Journal of the Belarusian State University. Law. 2021;(3):23–31.

20. Antopolsky A. B. International Activities in the Field of Language Technologies and Resources. Information and Innovations. 2021;16(2):5–22. (In Russ.). https://doi.org/10.31432/1994-2443-2021-16-2-5-22