论文

BadGCL:针对图持续学习系统的持久化后门攻击研究

BadGCL: Towards Persistent Backdoor Attack on Graph Continual Learning Systems

精选理由

做图神经网络或持续学习的同学该看看这篇。它证明了持续更新并不会冲掉后门,反而可能因为知识保留机制把攻击一直带着,实验跑了四个基准和两种协议,防御方案也得对照着改。

arXiv 论文提出 BadGCL,一个面向图持续学习(GCL)系统的模型无关持久化后门攻击框架。此前普遍认为持续更新会让后门随任务更替而失效,该工作指出 GCL 的知识保留机制反而会在中毒更新与干净任务信号兼容时把早期注入的后门一并保存下来。

正文

技术上,BadGCL 用梯度互信息目标学习拓扑感知的结构化触发器,把中毒梯度与干净任务的优化信号耦合,并配合传播一致性目标,让触发效果经 GNN 消息传递后在跨任务中保持稳定。论文给出理论分析,证明梯度兼容性与传播稳定性共同支撑后门在良性持续更新下长期存活。

实验覆盖 task-IL 与 class-IL 两种协议、四个 GCL 基准和多种受害系统,结果显示 BadGCL 在干净精度损失有限的情况下持续有效,明显优于图后门与持续学习后门基线,并对代表性防御保持鲁棒。对部署 GNN 持续学习系统的团队而言,训练早期的数据投毒风险需要重新评估。

原文 · arXiv cs.AI

BadGCL: Towards Persistent Backdoor Attack on Graph Continual Learning Systems

Backdoor attacks pose a persistent security threat to machine learning systems, yet their behavior under graph continual learning (GCL) remains underexplored. Conventional graph backdoor attacks often lose effectiveness as models are updated across successive tasks, creating the impression that continual updates naturally weaken malicious associations. This work reveals a contrasting vulnerability: knowledge-preservation mechanisms in GCL can also preserve early injected backdoors when the poisoned update is compatible with the clean-task signal being retained. For graph data, this risk is amplified by structural triggers, which are embedded into the message-passing paths of graph neural networks (GNNs) and can induce stable representation shifts across tasks. To expose this risk, we propose BadGCL, a model-agnostic framework for persistent backdoor attacks under a strict early-stage intervention setting. BadGCL learns topology-aware structural triggers using a gradient mutual-information objective that couples poisoned gradients with clean-task optimization signals, together with a propagation consistency objective that stabilizes trigger-induced effects through graph message passing. We provide theoretical analysis showing that gradient compatibility and propagation stability jointly support long-term backdoor persistence under benign continual updates. Extensive experiments across task-IL and class-IL protocols, four GCL benchmarks, and diverse victim systems show that BadGCL achieves persistent attack effectiveness with limited clean-utility degradation, substantially outperforming graph and continual-learning backdoor baselines and remaining robust against representative defenses.

产业坐标与信源核验

本条情报归属 AITOP「论文」核心追踪栏目。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:GNN、后门攻击、持续学习、图神经网络,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。

信源核验说明:本文原始线索采编自「arXiv cs.AI」(@Dong Li, Yufei Han, Xintao Wu, Xujiang Zhao, Wei Cheng, Chen Zhao),经 AITOP 多维评分管线结构化提炼。