Examining the Ethical Risks of Generative AI in Cybersecurity: An Experimental Study on Ethical, Gray Area and Unethical Usage Scenarios

Büşra Takgil

Makale detayları

Examining the Ethical Risks of Generative AI in Cybersecurity: An Experimental Study on Ethical, Gray Area and Unethical Usage Scenarios

Yazarlar:

Büşra Takgil

Yayıncı:

Duzce University

Ana Başlıklar:

Siber Güvenlik Politikaları | Yapay Zeka (Diğer)

Dergi:

Siber Güvenlik ve Dijital Ekonomi Vol.1 Issue 1, pp.1-9

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001353910

pISSN:

3108-3706

Anahtar kelimeler:

Özet:

 Generative AIGenAI systems, which are among the emerging artificial intelligence technologies, have started to play an important role in the field of cyber security for both defense and attack purposes. This study aims to systematically analyze the ethical risks of GenAI in cybersecurity. In the study, firstly, a literature review on the usage areas of GenAI in cybersecurity was conducted, then ethical, gray area and unethical usage scenarios were developed and applied experiments were carried out. Each scenario was tested separately using large language modelsLLM such as OpenAI ChatGPT and DeepSeek, and metrics such as accuracy rates, false positive rates and ethical risks were analyzed. The applied results have shown that GenAI-based systems can achieve high accuracy rates in early detection of security threats, but at the same time, they can cause serious ethical issues such as individual privacy violations, misleading content production, and malicious use. The findings of the study emphasize the need for stronger policy regulations, technical limitations, and ethical frameworks to ensure the safe and ethical use of Generative AI technologies in cybersecurity. The results also provide significant contributions to the academic literature and practitioners on how GenAI systems should be managed from both defensive and offensive perspectives.

 Generative AIGenAI systems, which are among the emerging artificial intelligence technologies, have started to play an important role in the field of cyber security for both defense and attack purposes. This study aims to systematically analyze the ethical risks of GenAI in cybersecurity. In the study, firstly, a literature review on the usage areas of GenAI in cybersecurity was conducted, then ethical, gray area and unethical usage scenarios were developed and applied experiments were carried out. Each scenario was tested separately using large language modelsLLM such as OpenAI ChatGPT and DeepSeek, and metrics such as accuracy rates, false positive rates and ethical risks were analyzed. The applied results have shown that GenAI-based systems can achieve high accuracy rates in early detection of security threats, but at the same time, they can cause serious ethical issues such as individual privacy violations, misleading content production, and malicious use. The findings of the study emphasize the need for stronger policy regulations, technical limitations, and ethical frameworks to ensure the safe and ethical use of Generative AI technologies in cybersecurity. The results also provide significant contributions to the academic literature and practitioners on how GenAI systems should be managed from both defensive and offensive perspectives.

 Generative AIGenAI systems, which are among the emerging artificial intelligence technologies, have started to play an important role in the field of cyber security for both defense and attack purposes. This study aims to systematically analyze the ethical risks of GenAI in cybersecurity. In the study, firstly, a literature review on the usage areas of GenAI in cybersecurity was conducted, then ethical, gray area and unethical usage scenarios were developed and applied experiments were carried out. Each scenario was tested separately using large language modelsLLM such as OpenAI ChatGPT and DeepSeek, and metrics such as accuracy rates, false positive rates and ethical risks were analyzed. The applied results have shown that GenAI-based systems can achieve high accuracy rates in early detection of security threats, but at the same time, they can cause serious ethical issues such as individual privacy violations, misleading content production, and malicious use. The findings of the study emphasize the need for stronger policy regulations, technical limitations, and ethical frameworks to ensure the safe and ethical use of Generative AI technologies in cybersecurity. The results also provide significant contributions to the academic literature and practitioners on how GenAI systems should be managed from both defensive and offensive perspectives.

 Gelişen yapay zekâ teknolojileri arasında yer alan Generative AI GenAI sistemleri, siber güvenlik alanında hem savunma hem de saldırı amaçlı olarak önemli bir rol oynamaya başlamıştır. Bu çalışma, GenAInin siber güvenlik alanındaki etik risklerini sistematik bir şekilde analiz etmeyi amaçlamaktadır. Çalışmada öncelikle GenAInin siber güvenlikteki kullanım alanlarına yönelik literatür taraması yapılmış, ardından etik, gri alan ve etik dışı kullanım senaryoları geliştirilerek uygulamalı deneyler gerçekleştirilmiştir. OpenAI ChatGPT ve DeepSeek gibi büyük dil modelleri LLM kullanılarak her senaryo ayrı ayrı test edilmiş; doğruluk oranları, yanlış pozitif oranları ve etik riskler gibi metrikler analiz edilmiştir. Uygulamalı sonuçlar, GenAI tabanlı sistemlerin, güvenlik tehditlerini erken tespit etmede yüksek doğruluk oranlarına ulaşabildiğini ancak aynı zamanda bireysel gizlilik ihlalleri, yanıltıcı içerik üretimi ve kötü amaçlı kullanım gibi ciddi etik sorunlar doğurabileceğini göstermiştir. Çalışmanın bulguları, Generative AI teknolojilerinin siber güvenlikte güvenli ve etik kullanımını sağlamak için daha güçlü politika düzenlemeleri, teknik sınırlamalar ve etik çerçevelerin geliştirilmesi gerektiğini vurgulamaktadır. Elde edilen sonuçlar ayrıca, hem savunma hem de saldırı perspektiflerinden GenAI sistemlerinin nasıl yönetilmesi gerektiğine dair akademik literatüre ve uygulayıcılara önemli katkılar sunmaktadır.

Makale detayları

Examining the Ethical Risks of Generative AI in Cybersecurity: An Experimental Study on Ethical, Gray Area and Unethical Usage Scenarios

Yazarlar:

Büşra Takgil

Yayıncı:

Duzce University

Ana Başlıklar:

Siber Güvenlik Politikaları | Yapay Zeka (Diğer)

Dergi:

Siber Güvenlik ve Dijital Ekonomi Vol.1 Issue 1, pp.1-9

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001353910

pISSN:

3108-3706

Generative AI

Cybersecurity

Ethical Risks

Gray Area Uses

Artificial Intelligence SecurityGenerative AI

Siber Güvenlik

Etik Riskler

Gri Alan Kullanımları

Yapay Zeka Güvenliği

Sayı

Makale detayları

Yazarlar: Büşra Takgil

Yayıncı: Duzce University

Ana Başlıklar: Siber Güvenlik Politikaları | Yapay Zeka (Diğer)

Dergi: Siber Güvenlik ve Dijital Ekonomi Vol.1 Issue 1, pp.1-9

Belge Tipi: Makale

Yayın Tarihi: 6/2025

Belge Dili: İngilizce

Document ID (Org No.): 2001353910

pISSN: 3108-3706

Sayı

Yazarlar:

Büşra Takgil

Yayıncı:

Duzce University

Ana Başlıklar:

Siber Güvenlik Politikaları | Yapay Zeka (Diğer)

Dergi:

Siber Güvenlik ve Dijital Ekonomi Vol.1 Issue 1, pp.1-9

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001353910

pISSN:

3108-3706