AI-Based Software Vulnerability Detection: A Systematic Literature Review

TL;DR

This paper systematically reviews AI-driven software vulnerability detection methods from 2018 to 2023, highlighting prevalent techniques, limitations, and future research directions in cybersecurity.

Contribution

It provides a comprehensive taxonomy of AI-based SVD techniques, analyzes current research trends, and identifies key challenges and emerging opportunities in the field.

Findings

91% of studies use AI-based methods

Graph-based models are most prevalent

Identified limitations include dataset quality and reproducibility

Abstract

Software vulnerabilities in source code pose serious cybersecurity risks, prompting a shift from traditional detection methods (e.g., static analysis, rule-based matching) to AI-driven approaches. This study presents a systematic review of software vulnerability detection (SVD) research from 2018 to 2023, offering a comprehensive taxonomy of techniques, feature representations, and embedding methods. Our analysis reveals that 91% of studies use AI-based methods, with graph-based models being the most prevalent. We identify key limitations, including dataset quality, reproducibility, and interpretability, and highlight emerging opportunities in underexplored techniques such as federated learning and quantum neural networks, providing a roadmap for future research.