Analyzing and Classifying the Poisoning Attacks in Federated Learning

DOI

10.2991/978-94-6239-693-7_105How to use a DOI?

Keywords

Poisoning Attacks; Data Poisoning; Model Poisoning; Byzantine Attacks; Secure Federated Learning

Abstract

Federated Learning (FL) is a potential distributed machine learning paradigm that protects user privacy by allowing collaborative model training without direct data sharing. Notwithstanding its benefits, FL is extremely susceptible to poisoning attacks, in which malevolent actors purposefully alter model updates or training data in order to impair performance or create backdoors. This study provides a thorough examination and categorization of poisoning assaults in federated learning. A decentralized machine learning system called collaborative learning allows several clients to work together to train a common global model without disclosing their local data. Although FL greatly enhances data governance and privacy protection, its scattered and partially trusted environment creates substantial security risks. One of the most critical threats is infecting where spiteful training data updates can corrupt the learning process. This research presents a step-by-step, in-depth analysis of destructive attacks in collaborative learning. We systematically describe the FL architecture, define a comprehensive threat model, classify poisoning attacks based on attack surface and adversarial objectives, and analyze their impact on model performance and reliability. Furthermore, we discuss existing defense mechanisms, provide result-oriented discussion based on experimental insights from literature, and conclude with key findings and future research directions.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Abdul Ahad
AU  - Mohammed Ali Shaik
PY  - 2026
DA  - 2026/06/16
TI  - Analyzing and Classifying the Poisoning Attacks in Federated Learning
BT  - Proceedings of the International Conference on Intelligent Systems for a Sustainable Future (ISSF 2026)
PB  - Atlantis Press
SP  - 1090
EP  - 1095
SN  - 2589-4919
UR  - https://doi.org/10.2991/978-94-6239-693-7_105
DO  - 10.2991/978-94-6239-693-7_105
ID  - Ahad2026
ER  -