Generative AI and Cybersecurity Education, Training, and Awareness: A Systematic Mapping Study
IEEE Access, cilt.14, ss.150806-150827, 2026 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Derleme
- Cilt numarası: 14
- Basım Tarihi: 2026
- Doi Numarası: 10.1109/access.2026.3738195
- Dergi Adı: IEEE Access
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, INSPEC, Directory of Open Access Journals
- Sayfa Sayıları: ss.150806-150827
- Anahtar Kelimeler: Cybersecurity education, cybersecurity training and awareness, generative AI, large language models
- Bursa Uludağ Üniversitesi Adresli: Evet
Özet
Large language models (LLMs) and generative artificial intelligence (AI) are increasingly explored in cybersecurity education, training, and awareness. These tools have been used for content generation, phishing awareness, cyber range scenario design, assessment support, tutoring, curriculum development, and hands-on learning activities. However, the literature remains scattered across educational settings, technical domains, learner groups, and tool types. This systematic mapping study searched IEEE Xplore, Taylor & Francis Online, Scopus, and Wiley Online Library on April 9, 2026. Of the 512 identified records, 112 were retained after title and abstract screening; 100 unique records underwent full-text eligibility assessment; and 60 peer-reviewed, English-language studies were included in the final synthesis. The review addresses three questions: the field's demographic landscape, the problems existing studies address, and the use of generative AI and LLMs in cybersecurity teaching and learning. The findings show that current studies primarily address limitations in traditional cybersecurity education, including static training materials, limited personalization, a lack of realistic scenarios, a shortage of adaptive feedback, and difficulty aligning learning activities with industry needs. Generative AI is commonly used to generate learning materials, support phishing and awareness training, assist with capture-the-flag (CTF) activities, create cyber range scenarios, provide tutoring support, and support assessment design and curriculum development. At the same time, the reviewed studies report concerns about hallucinations, factual accuracy, bias, overreliance, lack of empirical validation, and limited domain-specific datasets. The review indicates that generative AI has shown potential to support cybersecurity learning. Still, interpret the current evidence with caution because many studies evaluate content quality, usability, engagement, prototype performance, or short-term task outcomes rather than sustained learning, transfer to authentic cybersecurity tasks, or workforce competence.