Cross-Paradigm Graph Backdoor Attacks with Promptable Subgraph Triggers

TL;DR

This paper introduces CP-GBA, a novel method for creating transferable backdoor triggers in graph neural networks using promptable subgraph triggers, enhancing attack success across various paradigms.

Contribution

The paper develops a cross-paradigm attack framework employing graph prompt learning to generate transferable subgraph triggers, along with theoretical insights into GPL transferability.

Findings

Achieves state-of-the-art attack success rates across multiple datasets.

Demonstrates high transferability of triggers across different graph learning paradigms.

Maintains effectiveness under various defense scenarios.

Abstract

Graph Neural Networks(GNNs) are vulnerable to backdoor attacks, where adversaries implant malicious triggers to manipulate model predictions. Existing trigger generators are often simplistic in structure and overly reliant on specific features, confining them to a single graph learning paradigm, such as graph supervised learning, graph contrastive learning, or graph prompt learning. Such paradigm-specific designs lead to poor transferability across different learning frameworks, limiting attack success rates in general testing scenarios. To bridge this gap, we propose Cross-Paradigm Graph Backdoor Attacks with Promptable Subgraph Triggers(CP-GBA), which employs Graph Prompt Learning(GPL) to synthesize transferable subgraph triggers. Specifically, we first distill a compact yet expressive trigger set into a queryable repository, jointly optimizing for class-awareness, feature richness, and structural fidelity. Furthermore, we pioneer the theoretical exploration of GPL transferability under prompt-based objectives, ensuring robust generalization to diverse and unseen test-time paradigms. Extensive experiments across multiple real-world datasets and defense scenarios show that CP-GBA achieves state-of-the-art attack success rates.