SyncAI.news, a Varaisys broadcasting
Detecting Data Poisoning in Code Generation LLMs via Black-Box, Vulnerability-Oriented Scanning
SY

Shenao Yan, Shan Jin, Shimaa Ahmed, Sunpreet Singh Arora, Yiwei Cai, Yizhen Wang, Yuan Hong

· 1 min read

ResearcharXiv cs.AI

Detecting Data Poisoning in Code Generation LLMs via Black-Box, Vulnerability-Oriented Scanning

arXiv:2603.17174v2 Announce Type: replace-cross Abstract: Code generation large language models (LLMs) are increasingly integrated into modern software development workflows. Recent work has shown that these models are vulnerable to backdoor and poisoning attacks that induce the generation of insecure code, yet effective defenses remain limited. Existing scanning approaches rely on token-level generation consistency to invert attack targets, which is ineffective for source code where identical semantics can appear in diverse syntactic forms. We present CodeScan, the first black-box, vulnerability-specific scanning framework for auditing code generation LLMs, assuming that the defender specifies the target vulnerability classes and provides corresponding task-relevant prompts. CodeScan identifies attack targets by analyzing structural similarities across multiple generations conditioned on different clean prompts. It combines iterative divergence analysis with abstract syntax tree (AST)-based normalization to abstract away surface-level variation and unify semantically equivalent code, isolating structures that recur consistently across generations. CodeScan then applies LLM-based vulnerability analysis to determine whether the extracted structures contain security vulnerabilities and flags the model as compromised when such a structure is found. We evaluate CodeScan against four representative attacks under both backdoor and poisoning settings across three real-world vulnerability classes. Experiments on 117 models spanning three architectures and multiple model sizes demonstrate 97%+ detection accuracy with substantially lower false positives than prior methods.

Original source

This story was published by arXiv cs.AI and written by Shenao Yan, Shan Jin, Shimaa Ahmed, Sunpreet Singh Arora, Yiwei Cai, Yizhen Wang, Yuan Hong. SyncAI.news shows a preview; the complete article is on the publisher's site.

Read the full story on arxiv.org

Similar News