Evaluating and Improving the Robustness of Security Attack Detectors Generated by LLMs

TL;DR

This paper presents a novel approach combining Retrieval Augmented Generation and Self-Ranking to enhance the robustness and accuracy of security attack detectors generated by Large Language Models, significantly improving detection performance.

Contribution

It introduces an integrated pipeline using RAG and Self-Ranking to improve LLM-generated security detectors, addressing knowledge limitations and robustness challenges.

Findings

Up to 71%pt improvement in XSS detection F2-Score.

Up to 43%pt improvement in SQL injection detection F2-Score.

Significant overall enhancement in attack detection robustness.

Abstract

Large Language Models (LLMs) are increasingly used in software development to generate functions, such as attack detectors, that implement security requirements. A key challenge is ensuring the LLMs have enough knowledge to address specific security requirements, such as information about existing attacks. For this, we propose an approach integrating Retrieval Augmented Generation (RAG) and Self-Ranking into the LLM pipeline. RAG enhances the robustness of the output by incorporating external knowledge sources, while the Self-Ranking technique, inspired by the concept of Self-Consistency, generates multiple reasoning paths and creates ranks to select the most robust detector. Our extensive empirical study targets code generated by LLMs to detect two prevalent injection attacks in web security: Cross-Site Scripting (XSS) and SQL injection (SQLi). Results show a significant improvement in detection performance while employing RAG and Self-Ranking, with an increase of up to 71%pt (on average 37%pt) and up to 43%pt (on average 6%pt) in the F2-Score for XSS and SQLi detection, respectively.