Vulnerability Detection and Cyber Attack Simulation using Deep Learning
Date Issued
April 5, 2024
Type
Μεταπτυχιακή Διπλωματική Εργασία
Abstract
The integration of artificial intelligence (AI) in cybersecurity has ushered in a new era of advanced security measures, effectively tackling the complexities of cyber threats. Tools like Deep Exploit have showcased the potential of AI in penetration testing despite facing challenges. The development of AI-PenTest aims to overcome limitations observed in previous tools by focusing specifically on automating the processes an attacker would use to gain access to target machines. Utilizing a custom dataset and a pre-trained GPT-2 model, AI-PenTest demonstrates success in penetration testing, supporting the idea of AI in cybersecurity. The ongoing evolution of AI technology promises a new era in cybersecurity, enhancing the resilience of digital infrastructures against persistent threats.
Subjects