Evaluating Query Efficiency and Accuracy of Transfer Learning-based Model Extraction Attack in Federated Learning

TL;DR

This paper investigates the vulnerability of federated learning models to transfer learning-based model extraction attacks, demonstrating that attackers can effectively replicate models with high accuracy and fidelity, especially using pretrained models and limited queries.

Contribution

It introduces a transfer learning approach to model extraction attacks in federated learning and evaluates its effectiveness across multiple architectures and datasets.

Findings

Transfer learning enhances attack success with fewer queries.

Model fidelity correlates with query set size.

Pretrained models improve extraction accuracy.

Abstract

Federated Learning (FL) is a collaborative learning framework designed to protect client data, yet it remains highly vulnerable to Intellectual Property (IP) threats. Model extraction (ME) attacks pose a significant risk to Machine Learning as a Service (MLaaS) platforms, enabling attackers to replicate confidential models by querying black-box (without internal insight) APIs. Despite FL's privacy-preserving goals, its distributed nature makes it particularly susceptible to such attacks. This paper examines the vulnerability of FL-based victim models to two types of model extraction attacks. For various federated clients built under the NVFlare platform, we implemented ME attacks across two deep learning architectures and three image datasets. We evaluate the proposed ME attack performance using various metrics, including accuracy, fidelity, and KL divergence. The experiments show that for different FL clients, the accuracy and fidelity of the extracted model are closely related to the size of the attack query set. Additionally, we explore a transfer learning based approach where pretrained models serve as the starting point for the extraction process. The results indicate that the accuracy and fidelity of the fine-tuned pretrained extraction models are notably higher, particularly with smaller query sets, highlighting potential advantages for attackers.