Impact of Biased Data Injection on Model Integrity in Federated Learning

Manuel Lengl, Marc Benesch, Stefan Röhrl, Simon Schumann, Martin Knopp, Martin Knopp, Oliver Hayden, Klaus Diepold

2025

Abstract

Federated Learning (FL) has emerged as a promising solution in the medical domain to overcome challenges related to data privacy and learning efficiency. However, its federated nature exposes it to privacy attacks and model degradation risks posed by individual clients. The primary objective of this work is to analyze how different data biases (introduced by a single client) influence the overall model’s performance in a Cross-Silo FL environment and whether these biases can be exploited to extract information about other clients. We demonstrate, using two datasets, that bias injection can significantly affect model integrity, with the impact varying considerably across different datasets. Furthermore, we show that minimal effort is sufficient to infer the number of training samples contributed by other clients. Our findings highlight the critical need for robust data security mechanisms in FL, as even a single compromised client can pose serious risks to the entire system.

Paper Citation

in Harvard Style

Lengl M., Benesch M., Röhrl S., Schumann S., Knopp M., Hayden O. and Diepold K. (2025). Impact of Biased Data Injection on Model Integrity in Federated Learning. In Proceedings of the 18th International Joint Conference on Biomedical Engineering Systems and Technologies - Volume 1: BIOINFORMATICS; ISBN 978-989-758-731-3, SciTePress, pages 432-443. DOI: 10.5220/0013153800003911

in Bibtex Style

@conference{bioinformatics25,
author={Manuel Lengl and Marc Benesch and Stefan Röhrl and Simon Schumann and Martin Knopp and Oliver Hayden and Klaus Diepold},
title={Impact of Biased Data Injection on Model Integrity in Federated Learning},
booktitle={Proceedings of the 18th International Joint Conference on Biomedical Engineering Systems and Technologies - Volume 1: BIOINFORMATICS},
year={2025},
pages={432-443},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013153800003911},
isbn={978-989-758-731-3},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 18th International Joint Conference on Biomedical Engineering Systems and Technologies - Volume 1: BIOINFORMATICS
TI - Impact of Biased Data Injection on Model Integrity in Federated Learning
SN - 978-989-758-731-3
AU - Lengl M.
AU - Benesch M.
AU - Röhrl S.
AU - Schumann S.
AU - Knopp M.
AU - Hayden O.
AU - Diepold K.
PY - 2025
SP - 432
EP - 443
DO - 10.5220/0013153800003911
PB - SciTePress