Generative Neural Networks as a Tool for Web Applications Penetration Testing

• Published in: 2023 Communication and Information Technologies (KIT) pp. 1 - 5
• Main Authors: Gallus, Petr, Stepanek, Marcel, Racil, Tomas, Frantis, Petr
• Format: Conference Proceeding
• Language: English
• Published: IEEE 11-10-2023
• Subjects: Chatbots; ChatGPT; Cybersecurity; Generative Neural Networks; Industries; Information and communication technology; Knowledge engineering; Neural networks; OpenAI; Organizations; Penetration testing; Research and development; Web application; Wordpress
• DOI: 10.1109/KIT59097.2023.10297109

The scientific paper delves into the potential of generative neural networks as a powerful tool for web application penetration testing. By leveraging the capabilities of these networks, we aim to augment traditional testing methodologies and advance the field of vulnerability detection. In the second section, the paper provides an overview of OpenAI, a leading organization at the forefront of artificial intelligence research and development. OpenAI has contributed significantly to the field of natural language processing and has developed advanced models like ChatGPT, which have the potential to revolutionize various industries, including cybersecurity. It explores the underlying technology behind ChatGPT and discusses its implications for the field of web application penetration testing. Third section focuses on present the details of the experimental setup. A series of three experiments was conducted to evaluate the effectiveness of generative neural networks, specifically ChatGPT, in web application penetration testing. Through these experiments, its aim was to demonstrate the practical application of generative neural networks in identifying and exploiting web-based security vulnerabilities. In the fourth section, the results obtained from the experiments are presented. Parts of the experiment were categorized into three sub-results, each highlighting a specific aspect of vulnerability detection. The main intention was to highlight the potential of generative neural networks as an innovative and effective tool for web application penetration testing. In conclusion, the paper showcases the advancements made possible by generative neural networks in the domain of web application penetration testing. By automating certain aspects of the testing process and enhancing vulnerability detection, these networks hold immense promise for improving the overall security posture of web-based systems. The findings presented in this paper contribute to the growing body of knowledge in the field and open up new avenues for further research and development in this critical area of cybersecurity.