Fuzzing, a widely-used technique for bug detection, has seen advancements through Large Language Models (LLMs). Despite their potential, LLMs face specific challenges in fuzzing. In this paper, we identified five major challenges of LLM-assisted fuzzing. To support our findings, we revisited the most recent papers from top-tier conferences, confirming that these challenges are widespread. As a remedy, we propose some actionable recommendations to help improve applying LLM in Fuzzing and conduct preliminary evaluations on DBMS fuzzing. The results demonstrate that our recommendations effectively address the identified challenges.
@article{arxiv.2404.16297,
title = {When Fuzzing Meets LLMs: Challenges and Opportunities},
author = {Yu Jiang and Jie Liang and Fuchen Ma and Yuanliang Chen and Chijin Zhou and Yuheng Shen and Zhiyong Wu and Jingzhou Fu and Mingzhe Wang and ShanShan Li and Quan Zhang},
journal= {arXiv preprint arXiv:2404.16297},
year = {2024}
}