arXiv 论文提出自然代码变换可干扰智能体漏洞分析
Natural Code Transformation Confuses Agents in Vulnerability Analysis
AI 导读
芝加哥大学等研究者发现,AI 智能体在功能与漏洞相同但代码形式不同的真实提交上表现差异巨大。论文从真实开发模式中挖掘自然且保持功能的代码变换,实验显示单一变换即可显著降低 Claude Mythos 5.2 配合 Claude Code 在漏洞发现中的成功率,并研究了变换组合与自适应攻击者的逆向尝试。
正文 · 原文
Abstract:Having demonstrated a strong potential for automating vulnerability discovery, AI agents can also be used to attack software systems at scale. Today's software systems need mechanisms to level the playing field by slowing down potential attacks, giving them time to patch vulnerabilities and strengthen software. Along this line, our study of agent trajectories in vulnerability analysis shows a surprising finding: agents produce dramatically different results across real-world code commits that carry the same functionality and the same vulnerability. This phenomenon appears consistently in multiple code domains and vulnerability categories.
It motivates us to study natural and functionality-preserving code transformations derived from real-world software development, and their potential as a defense against attacker agents. We build a corpus of natural transformations by mining recurring development patterns from real-world code commits, and apply them to evaluate impact on agentic vulnerability analysis. Our results show that even a single form of transformation can effectively "confuse" frontier agents, e.g., Claude Mythos 5.2 with Claude Code, reducing successes in vulnerability discovery. We also study different agent failures, whether transformations can be composed to provide stronger protection, and adaptive attackers who seek to reverse them. Overall, our study of natural code transformation significantly expands the research space of using code transformations to defend against malicious agents.
| Subjects: | Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.14350 [cs.CR] |
| (or arXiv:2610.14350v3 [cs.CR] for this version) | |
| https://doi.org/10.48550/arXiv.2610.14350 arXiv-issued DOI via DataCite |
Submission history
From: Jie Zhu [view email]
[v1]
Fri, 9 Oct 2026 21:55:41 UTC (186 KB)
[v2]
Wed, 14 Oct 2026 13:20:51 UTC (187 KB)
[v3]
Tue, 17 Nov 2026 03:37:35 UTC (252 KB)
来源:arXiv cs.AI · arxiv.org