Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper investigates the applicability of open-source LLM frameworks, including both large-scale and lightweight models, for automating penetration testing tasks relevant to commercial security assessments. The study identifies both the potential and limitations of these frameworks in addressing fundamental challenges in penetration testing. The authors propose a practical approach to overcome key limitations and demonstrate the potential of LLM-based frameworks in real-world penetration testing scenarios.
LLMs show promise in automating penetration testing, but face practical limitations that this paper addresses with a proposed approach for real-world application.
The rapid development of artificial intelligence technologies, particularly Large Language Models, has a growing impact on the field of information security. One of the promising directions of their application is the automation of penetration testing - a controlled simulation of intruder actions to identify vulnerabilities in protected information systems.This study investigates the practical applicability of existing open-source frameworks based on Large Language Models, both large-scale and lightweight models, within the context of penetration testing tasks that are critical to commercial security assessments.The results demonstrated the potential application and constraints of these frameworks, addressing fundamental challenges in penetration testing and outline possible solutions for overcoming the identified limitations. Based on our results, we propose a practical approach to address key limitations and showcase the potential of Large Language Models based frameworks in real-world penetration testing.