Articles

Cyber threats of LLM implementation in state and municipal information systems

Received: 03.12.2025 Published: 15.11.2025

EDN: JKODWP

Full text

Abstract

The article discusses three main types of prompt injections: direct, indirect, and those aimed at generating malicious instructions. In the context of LLM application in the public corporate sector, the authors propose a systematic approach to their analysis. The study examines the risks associated with employing LLMs to support electronic document management, prepare official documents, and automate the processing of citizens’ requests. The results obtained indicate the need to develop specialized protection methods and regulations for the use of LLM. They also highlight the need to introduce mechanisms for validating model conclusions and to limit the powers of agent systems.

Keywords

Received: 3 December 2025

Agencies

The study was conducted as part of the state-funded project SPb FRC RAS No. FFZF-2025-0006.

How to cite

Для цитирования:

Вяткин А.А., Мурадян Д.С., Попцов А.В., и др. Киберугрозы внедрения LLM в государственные и муниципальные информационные системы // Государственная служба. 2025. Т. 27, № 6. С. 16–25.

For citation:

Vyatkin A.A., Muradyan D.S., Poptsov A.V., et al. Cyber threats of LLM implementation in state and municipal information systems // Gosudarstvennaya Sluzhba. 2025. Vol. 27, no. 6. P. 16–25.

Share

How to Cite

Vyatkin A.A., Muradyan D.S., Poptsov A.V., et al. Cyber threats of LLM implementation in state and municipal information systems // Gosudarstvennaya Sluzhba. 2025. Vol. 27, no. 6. P. 16–25.

Коды статьи

  • EDN JKODWP
  • Тип статьи (РИНЦ) Научная статья (Научная статья)

References

  1. Селиверстова А.Д. Возможности и риски, этические проблемы использования искусственного интеллекта в публичном управлении. Гуманитарные, социально-экономические и общественные науки. 2024. № 3. С. 56–60. https://doi.org/10.23672/SAE.2024.88.31.023
  2. Семин А.Н., Гусманов Р.У., Стовба Е.В., Мешкова Н.Г. Применение технологий искусственного интеллекта в органах государственной власти: вызовы и перспективы. ЭТАП: экономическая теория, анализ, практика. 2025. № 5. С. 33–44. https://doi.org/10.24412/2071-6435-2025-5-33-44.
  3. Талапина Э.В. Искусственный интеллект и правовые экспертизы в государственном управлении. Вестник Санкт-Петербургского университета. Серия: Право. 2021. № 4. С. 865–881.
  4. Ahmad N., Anshari M., Hamdan M., Ali E. Ethics and Public Trust in AI Governance: A Literature Review. International Journal of Law Government and Communication. 2024. Vol. 9. No. 37. P. 296–305. https://doi.org/10.35631/IJLGC.937025. In English
  5. Gulyamov S., Gulyamov S., Rodionov A., Khursanov R., Mekhmonov K., Babaev D., Rakhimjonov A. Prompt Injection Attacks in Large Language Models and AI Agent Systems: A Comprehensive Review of Vulnerabilities, Attack Vectors, and Defense Mechanisms. Preprints. 2025. https://doi.org/10.20944/preprints202511.0088.v1. https://www.preprints.org/frontend/manuscript/40097b1a4d5fc29a26a1088e674be312/download_pub. In English
  6. Heverin T., Cohen E. Evaluating the Effectiveness of Psychological Prompt Injection Attacks on Large Language Models for Social Engineering Artifact Generation. European Conference on Cyber Warfare and Security. 2025. Vol. 24. No. 1. P. 879–883. https://doi.org/10.34190/eccws.24.1.3515. In English
  7. Huang L., Yu W., Ma W., Zhong W., Feng Z., Wang H., Chen Q., Peng W., Feng X., Qin B., Liu T. A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questions. ACM Transactions on Information Systems. 2025. Vol. 43. No. 2. P. 1–55. https://doi.org/10.1145/3703155. In English
  8. Kwon H., Pak W. Text-Based Prompt Injection Attack Using Mathematical Functions in Modern Large Language Models. Electronics. 2024. Vol. 13. No. 24. Article 5008. https://doi.org/10.3390/electronics13245008. In English
  9. Mathew E.S. Enhancing Security in Large Language Models: A Comprehensive Review of Prompt Injection Attacks and Defenses. Journal on Artificial Intelligence. 2025. Vol. 7. P. 347–363. https://doi.org/10.32604/jai.2025.069841. In English
  10. Şaşal S., Can Ö. Prompt Injection Attacks on Large Language Models: Multi-Model Security Analysis with Categorized Attack Types // Proceedings of the 17th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2025). Vol. 1: KDIR. 2025. P. 517–524. https://doi.org/10.5220/0013838400004000. In English
  11. Yang H., Nawi H.S.A., Zhang Y. Artificial Intelligence and Large Language Models in Government Document Management. Journal of Logistics, Informatics and Service Science. 2025. Vol. 12. No. 4. P. 129–145. In English
  12. Yi J., Xie Y., Zhu B., Kiciman E., Sun G., Xie X., Wu F. Benchmarking and Defending against Indirect Prompt Injection Attacks on Large Language Models // Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining. 2025. P. 1809–1820. https://doi.org/10.1145/3690624.3709179. In English
  13. Seliverstova A.D. Opportunities and risks, ethical problems of using artificial intelligence in public administration. Gumanitarnyye, sotsialno-ekonomicheskiye i obshchestvennyye nauki. 2024. No. 3. P. 56–60. https://doi.org/10.23672/SAE.2024.88.31.023. In Russian
  14. Semin A.N., Gusmanov R.U., Stovba E.V., Meshkova N.G. Application of artificial intelligence technologies in public authorities: challenges and prospects. ETAP: ekonomicheskaya teoriya, analiz, praktika. 2025. No. 5. P. 33–44. https://doi.org/10.24412/2071-6435-2025-5-33-44. In Russian
  15. Talapina E.V. Artificial intelligence and legal expertise in public administration. Vestnik Sankt-Peterburgskogo universiteta. Seriya: Pravo. 2021. No. 4. P. 865–881. In Russian