10-tom, 1-son, b. 457–462, Nashr: 2026-10-14

SUN’IY INTELLEKT VOSITALARI TALABA FAOLIYATIDA KELTIRIB CHIQARADIGAN AXBOROT TAHDIDLARINING PEDAGOGIK TASNIFI

ПЕДАГОГИЧЕСКАЯ КЛАССИФИКАЦИЯ ИНФОРМАЦИОННЫХ УГРОЗ, ВОЗНИКАЮЩИХ В ДЕЯТЕЛЬНОСТИ СТУДЕНТА ПРИ ИСПОЛЬЗОВАНИИ СРЕДСТВ ИСКУССТВЕННОГО ИНТЕЛЛЕКТА

PEDAGOGICAL CLASSIFICATION OF INFORMATION THREATS ARISING FROM THE USE OF ARTIFICIAL INTELLIGENCE TOOLS IN STUDENT ACTIVITY

• Mustafoyev Shahzod Mamanazar o‘g‘li — Buxoro Xalqaro Universiteti mustaqil izlanuvchisi · ORCID: https://orcid.org/0009-0007-8584-1237

To'liq matn (PDF)

Annotatsiya

Maqolada oliy ta’lim talabalari generativ sun’iy intellekt vositalari bilan kundalik ishlaganda duch keladigan axborot tahdidlarini pedagogik maqsadda tasniflash masalasi ko‘rib chiqilgan. Uchta oliy ta’lim muassasasida 586 talaba o‘rtasida so‘rov o‘tkazildi. Respondentlarning 71,3 foizi shaxsiy ma’lumotlarini sun’iy intellektga kiritgan, 64,7 foizi olingan natijani tekshirmay ishlatgan, 43,6 foizi esa rasmiy xizmat o‘rniga uchinchi shaxslar yaratgan Telegram-botlardan foydalanadi. OWASP, NIST va boshqa mavjud tasniflar muhandislar uchun tuzilgan, shu bois ularni to‘g‘ridan-to‘g‘ri o‘quv mazmuniga aylantirib bo‘lmaydi. Muallif tahdidlarni talabaning sun’iy intellekt bilan muloqoti bosqichlariga qarab to‘rt guruhga ajratadi: kiruvchi ma’lumotlar, chiquvchi natija, kirish va hisob qaydnomasi hamda kognitiv-xulqiy tahdidlar. Har bir guruh uchun shakllantirilishi kerak bo‘lgan himoya ko‘nikmasi belgilangan. Tasnifdan o‘quv mazmunini tanlash va keyslar tuzishda qanday foydalanish mumkinligi misollarda ko‘rsatilgan.

Аннотация

В статье рассматривается задача педагогической классификации информационных угроз, возникающих при повседневном взаимодействии студентов вузов с генеративными средствами искусственного интеллекта. Приводятся результаты опроса 586 студентов трёх вузов: 71,3% вводили в ИИ персональные данные, 64,7% использовали ответ без проверки, 43,6% обращаются к ИИ через сторонние Telegram-боты. Показано, что существующие классификации (OWASP, NIST и др.) построены с инженерной точки зрения и не могут быть напрямую переведены в учебное содержание. Предложено делить угрозы по этапам акта взаимодействия студента с ИИ на четыре группы – входные данные, выходной результат, доступ и учётная запись, когнитивно-поведенческие угрозы – с указанием формируемого защитного умения для каждой группы. На примерах показано применение классификации при отборе содержания и проектировании кейсов.

Abstract

The paper addresses the task of classifying, for pedagogical purposes, the information threats that arise in students’ everyday interaction with generative AI tools. Results of a survey of 586 students from three universities are reported: 71.3% had entered personal data into AI tools, 64.7% had used AI output without verification, and 43.6% access AI through third-party Telegram bots rather than official services. It is shown that existing classifications (OWASP, NIST and others) are built from an engineering standpoint and cannot be translated directly into curriculum content. The author proposes grouping threats by the stages of a student’s interaction act with AI into four categories – input data, output result, access and account, cognitive-behavioural threats – and specifies the protective skill to be developed for each. The use of the classification in content selection and case design is illustrated with examples.

Kalit so'zlar: sun’iy intellekt, axborot tahdidlari, pedagogik tasnif, gallyutsinatsiya, prompt-inyeksiya, deepfake, shaxsga doir ma’lumotlar, talabalar, xavfsiz xulq-atvor

Foydalanilgan adabiyotlar

1. OWASP Top 10 for Large Language Model Applications. Version 1.1. – OWASP Foundation, 2023. – 33 p.
2. National Institute of Standards and Technology. Artificial Intelligence Risk Management Framework (AI RMF 1.0). – Gaithersburg: NIST, 2023. – 48 p.
3. Weidinger L., Mellor J., Rauh M. et al. Ethical and social risks of harm from Language Models. – arXiv:2112.04359, 2021. – 64 p.
4. ENISA Threat Landscape 2024. – Athens: European Union Agency for Cybersecurity, 2024. – 138 p.
5. O‘zbekiston Respublikasining "Shaxsga doir ma’lumotlar to‘g‘risida"gi Qonuni. 2019-yil 2-iyul, O‘RQ-547-son.
6. Gurman M. Samsung bans staff’s AI use after spotting ChatGPT data leak // Bloomberg. – 02.05.2023.
7. Ji Z., Lee N., Frieske R. et al. Survey of Hallucination in Natural Language Generation // ACM Computing Surveys. – 2023. – Vol. 55, № 12. – P. 1–38.
8. Weiser B. Here’s what happens when your lawyer uses ChatGPT // The New York Times. – 27.05.2023.
9. Greshake K., Abdelnabi S., Mishra S., Endres C., Holz T., Fritz M. Not what you’ve signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection // Proceedings of the 16th ACM Workshop on Artificial Intelligence and Security. – New York: ACM, 2023. – P. 79–90.
10. Tal N. "FakeGPT": New variant of fake-ChatGPT Chrome extension stealing Facebook ad accounts // Guardio Labs. – 09.03.2023.
11. Parasuraman R., Manzey D.H. Complacency and Bias in Human Use of Automation: An Attentional Integration // Human Factors. – 2010. – Vol. 52, № 3. – P. 381–410.
12. Westerlund M. The Emergence of Deepfake Technology: A Review // Technology Innovation Management Review. – 2019. – Vol. 9, № 11. – P. 39–52.
13. Chen H., Magramo K. Finance worker pays out $25 million after video call with deepfake "chief financial officer" // CNN. – 04.02.2024.
14. Kasneci E., Sessler K., Küchemann S. et al. ChatGPT for good? On opportunities and challenges of large language models for education // Learning and Individual Differences. – 2023. – Vol. 103. – Article 102274.
15. Vuorikari R., Kluzer S., Punie Y. DigComp 2.2: The Digital Competence Framework for Citizens. – Luxembourg: Publications Office of the European Union, 2022. – 133 p.