Trustworthy Graph Neural Networks: Aspects, Methods and Trends

TL;DR

This survey articulates a comprehensive roadmap for trustworthy GNNs by organizing six core aspects—robustness, explainability, privacy, fairness, accountability, and environmental well-being—and detailing their concepts, methodologies, and practical implications. It introduces an open framework that maps trustworthy AI principles to GNN-specific contexts and emphasizes cross-aspect interactions to guide design decisions. Through fine-grained taxonomies and cross-cutting discussions, the paper provides actionable guidance for defense, explanation, privacy, fairness, and efficiency in real-world GNN deployments. By outlining trends, datasets, and industrialisation pathways, it aims to accelerate the development of robust, transparent, and responsible graph learning systems. The work underscores the importance of model-agnostic approaches, ecosystem tooling, and scalable evaluations to translate trustworthy GNNs from theory to practice.

Abstract

Graph neural networks (GNNs) have emerged as a series of competent graph learning methods for diverse real-world scenarios, ranging from daily applications like recommendation systems and question answering to cutting-edge technologies such as drug discovery in life sciences and n-body simulation in astrophysics. However, task performance is not the only requirement for GNNs. Performance-oriented GNNs have exhibited potential adverse effects like vulnerability to adversarial attacks, unexplainable discrimination against disadvantaged groups, or excessive resource consumption in edge computing environments. To avoid these unintentional harms, it is necessary to build competent GNNs characterised by trustworthiness. To this end, we propose a comprehensive roadmap to build trustworthy GNNs from the view of the various computing technologies involved. In this survey, we introduce basic concepts and comprehensively summarise existing efforts for trustworthy GNNs from six aspects, including robustness, explainability, privacy, fairness, accountability, and environmental well-being. Additionally, we highlight the intricate cross-aspect relations between the above six aspects of trustworthy GNNs. Finally, we present a thorough overview of trending directions for facilitating the research and industrialisation of trustworthy GNNs.

Figures (4)

• Figure 1: The Open Framework of Trustworthy GNNs and Overview of Methodology Categorisation. Trustworthy GNNs are competent GNNs that incorporate aspects of trustworthiness, including robustness, explainability, privacy, fairness, accountability, well-being, and other trust-oriented aspects in the context of GNNs. For each aspect of trustworthy GNNs, we comprehensively summarise typical methods and provide methodology categorisation, as will be shown in the following sections.
• Figure 2: Process of different adversarial attacks against GNNs. Training and inference without adversarial perturbations can produce a clean GNN model and correct prediction results (as shown in the second column dotted box). Meanwhile, different perturbations to the clean graph are added during either a training or inference phases to produce evasion attacks (the first column dotted box), poisoning attacks (the third column dotted box), and backdoor attacks (the fourth column dotted box), which lead to different kinds of misclassifications ("Incorrect Results" box at the bottom).
• Figure 3: An overview of methodologies for improving the interpretability and explainability of GNNs. Self-interpretable GNNs (green background) mainly include contribution estimation, introduction of interpretable module, embedding prototype learning, and rationale generation methods. Explainers of GNNs (grey background) include gradient/feature-based methods, perturbation-based methods, surrogate methods, decomposition methods, generation methods, and other methods.
• Figure 4: The Unfairness Cycle in Systems with GNNs. The bias circulates between graph data, graph neural networks and users.