AI-Powered Cyberattacks: A Comprehensive Review and Analysis of Emerging Threats
[ 1 ] Szkoła Doktorska, Akademia Sztuki Wojennej | [ SzD ] doktorant ze Szkoły Doktorskiej
2025
artykuł naukowy
angielski
- AI-powered cyberattacks
- Offensive AI
- Cybersecurity
- Deepfake attacks
- Phishing automation
- AI-generated malware
- Cyberbezpieczeństwo
- Deepfake
- Phishing
- Złośliwe oprogramowanie
- Sztuczna inteligencja
EN The growing role of artificial intelligence (AI) in cybersecurity presents both opportunities and challenges. While AI strengthens detection and defense systems, it also enables cybercriminals to launch sophisticated, adaptive, and large-scale attacks.This paper reviews keyAI-powered cyber threats, including deepfake decep-tion, AI-driven phishing campaigns, automated vulnerability exploitation, and the autonomous generation of malware. It also examines emerging techniques such as prompt injection, model stealing, and data poisoning, which compromise the integrity of AI-based systems. In response, the study explores advanced defense strate-gies such as explainable AI (XAI), multimodal detection models, content watermarking, and AI-enhanced honeypots. It further considers regulatory frameworks and ethical concerns surrounding dual-use AI. By ana-lyzing current research and real-world incidents, this study supports a deeper understanding of evolving threats and outlines recommendations for proactive and collaborative cybersecurity efforts in the age of AI.
55 - 69
CC BY-NC-ND (uznanie autorstwa - użycie niekomercyjne - bez utworów zależnych)
otwarte czasopismo
ostateczna wersja opublikowana
25.09.2025
5