Stealthy Yet Effective: Distribution-Preserving Backdoor Attacks on Graph Classification
This addresses security vulnerabilities in graph neural networks for applications like social network analysis or bioinformatics, but it is incremental as it builds on existing backdoor attack methods by focusing on stealth improvements.
The paper tackles the problem of backdoor attacks on graph classification by identifying that existing methods are easily detectable due to structural and semantic anomalies, and proposes DPSBA, a clean-label framework that learns in-distribution triggers to improve stealth. The result shows DPSBA achieves a superior balance between attack success and detectability compared to state-of-the-art baselines, with extensive experiments validating its effectiveness.
Graph Neural Networks (GNNs) have demonstrated strong performance across tasks such as node classification, link prediction, and graph classification, but remain vulnerable to backdoor attacks that implant imperceptible triggers during training to control predictions. While node-level attacks exploit local message passing, graph-level attacks face the harder challenge of manipulating global representations while maintaining stealth. We identify two main sources of anomaly in existing graph classification backdoor methods: structural deviation from rare subgraph triggers and semantic deviation caused by label flipping, both of which make poisoned graphs easily detectable by anomaly detection models. To address this, we propose DPSBA, a clean-label backdoor framework that learns in-distribution triggers via adversarial training guided by anomaly-aware discriminators. DPSBA effectively suppresses both structural and semantic anomalies, achieving high attack success while significantly improving stealth. Extensive experiments on real-world datasets validate that DPSBA achieves a superior balance between effectiveness and detectability compared to state-of-the-art baselines.