CryptGNN: Enabling Secure Inference for Graph Neural Networks

2025/09/11 by Sen, Pritam, Ma, Yao, Borcea, Cristian
#Cryptography and Security (cs.CR) #FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · doi:10.48550/arxiv.2509.09107

Abstract

We present CryptGNN, a secure and effective inference solution for third-party graph neural network (GNN) models in the cloud, which are accessed by clients as ML as a service (MLaaS). The main novelty of CryptGNN is its secure message passing and feature transformation layers using distributed secure multi-party computation (SMPC) techniques. CryptGNN protects the client's input data and graph structure from the cloud provider and the third-party model owner, and it protects the model parameters from the cloud provider and the clients. CryptGNN works with any number of SMPC parties, does not require a trusted server, and is provably secure even if P-1 out of P parties in the cloud collude. Theoretical analysis and empirical experiments demonstrate the security and efficiency of CryptGNN.

Citations

• CrypTen: Secure Multi-Party Computation Meets Machine Learning
• Universal Representation for Code
• Efficient Privacy-Preserving Computation Based on Additive Secret Sharing
• Understanding Isomorphism Bias in Graph Data Sets
• CrypTFlow: Secure TensorFlow Inference
• XONN: XNOR-based Oblivious Deep Neural Network Inference
• How Powerful are Graph Neural Networks?
• Path ORAM
• Gazelle: A Low Latency Framework for Secure Neural Network Inference
• Chameleon: A Hybrid Secure Computation Framework for Machine Learning\n Applications
• Inductive Representation Learning on Large Graphs
• Revisiting Semi-Supervised Learning with Graph Embeddings
• Convolutional Networks on Graphs for Learning Molecular Fingerprints
• How to share a secret

Related

• SecGNN: Privacy-Preserving Graph Neural Network Training and Inference as a Cloud Service
• PrivGNN: High-Performance Secure Inference for Cryptographic Graph Neural Networks
• Graph in the Vault: Protecting Edge GNN Inference with Trusted Execution Environment
• Panther: A Cost-Effective Privacy-Preserving Framework for GNN Training and Inference Services in Cloud Environments
• DESIGN: Encrypted GNN Inference via Server-Side Input Graph Pruning