Paper 2025/2296
SoK: Verifiable Federated Learning
Abstract
Federated Learning (FL) is an advancement in Machine Learning motivated by the need to preserve the privacy of the data used to train models. While it effectively addresses this issue, the multi-participant paradigm on which it is based introduces several challenges. Among these are the risks that participating entities may behave dishonestly and fail to perform their tasks correctly. Moreover, due to the distributed nature of the architecture, attacks such as Sybil and collusion are possible. Recently, with advances in Verifiable Computation (VC) and Zero-Knowledge Proofs (ZKP), researchers have begun exploring how to apply these technologies to Federated Learning aiming to mitigate such problems. In this Systematization of Knowledge, we analyze the first, very recent works that attempt to integrate verifiability features into classical FL tasks, comparing their approaches and highlighting what is achievable with the current state of VC methods.
Metadata
- Available format(s)
-
PDF
- Category
- Applications
- Publication info
- Preprint.
- Keywords
- Machine LearningMLFederated LearningFLArtificial IntelligenceAIZero KnowledgeZKZKPVerifiability
- Contact author(s)
-
francesco bruschi @ polimi it
marco esposito @ polimi it
tomgag @ horizenlabs io
arizzini @ horizenlabs io - History
- 2025-12-22: approved
- 2025-12-20: received
- See all versions
- Short URL
- https://ia.cr/2025/2296
- License
-
CC BY-NC-ND
BibTeX
@misc{cryptoeprint:2025/2296,
author = {Francesco Bruschi and Marco Esposito and Tommaso Gagliardoni and Andrea Rizzini},
title = {{SoK}: Verifiable Federated Learning},
howpublished = {Cryptology {ePrint} Archive, Paper 2025/2296},
year = {2025},
url = {https://eprint.iacr.org/2025/2296}
}
