cs.AIOct 28, 2026

Governance-Aware Federated Learning for Multi-Agent AI Systems: A Policy-Driven Framework for Trustworthy Collaborative AI

Authors: Kiranmayee Janardhan, Karthick S, Madhusmita Panda, Bharathi Vokkaliga Ganesh, Syed Ahmed

Organizations: Responsible AI Office, Infosys Limited, Infosys Avenue, 3rd Cross, Electronic City, Hosur Road, Bengaluru – 560100, Karnataka, India

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Federated Learning (FL) enables collaborative model training without exchanging raw data, but conventional FL provides limited runtime governance, evidence verification, fairness assessment, explainability monitoring, and auditability. This paper proposes Governance-Aware Federated Learning (GAFL), which combines independent governance-evidence verification, temporal trust, fairness and explanation assessment, anomaly detection, policy enforcement, normalized aggregation, and tamper-evident audit logging. Client-reported governance values are authenticated, bound to their submitted model updates, and independently recomputed before aggregation. A cross-silo prototype with 10 participants was evaluated on the Adult Income dataset under honest, benign non-IID, biased-data, malicious-update, non-compliant, falsified-evidence, and mixed-threat scenarios. Four methods were compared across five random seeds: FedAvg, FedAvg with anomaly filtering, GAFL without independent verification, and complete GAFL. Complete GAFL intervened on all affected participants at the final round across the evaluated risk scenarios, including identifying 100% of falsified-evidence participants, whereas GAFL without verification identified none of the falsified-evidence participants. Final macro-F1 ranged from 0.7520 to 0.7733 across GAFL scenarios. Complete GAFL required 173.67 plus or minus 26.94 ms of governance processing per round, compared with 3.82 plus or minus 0.79 ms for FedAvg. A strict demographic-parity threshold also caused frequent interventions for honest participants, demonstrating a significant fairness-policy and utility trade-off.

More in this topic

Federated Learning

Federated learning trains shared models across data holders without requiring them to pool their raw data.

See all 961 papers

Explore similar work

CardsList
Oct 21, 2026

Extending FAIR Principles with Observability for Trustworthy Federated Learning

Federated learning (FL) keeps data at its source, but locality alone does not make a federation trustworthy. Without sufficient observability, an aggregator cannot determine whether the client service was healthy, the local data pipeline was trustworthy, or contractual constraints were properly monitored during training. We take the position that trustworthy FL needs an evidence plane in addition to its model-update process. Each participant continuously summarizes local infrastructure, data into minimal, signed trust evidence associated with each FL round, while raw data, logs remain local. A local TinyLLM may assist in interpreting observability signals, deriving contextual evidence, and providing natural-language explanations to operators. We instantiate this vision through FAIR-O, an observability extension of FAIR principle. Rather than benchmarking FL accuracy, a three-level trust model has been proposed and a two-participant Python prototype demonstrates evidence generation and trust assessment. We further discuss future integration of FAIR-O Trust Evidence into participant selection, evidence weighting, and federated aggregation.

Hangli Ge, Noboru Koshizuka

Interfaculty Initiative in Information Studies; The University of Tokyo

Trust-Aware Federated LearningFederated Learning
AskPDF
Oct 2, 2026

PoCoFL: POlicy-COmpliant Federated Learning

Federated Learning (FL) is a privacy-oriented learning paradigm that enables collaborative model training while keeping training data local to participating clients. However, it does not guarantee that clients submit policy-compliant contributions or that aggregators process admitted contributions correctly. Existing verifiable FL systems tailor validation rules to specific FL settings, learning workflows, and cryptographic constructions, limiting their applicability across network topologies, participant roles, and aggregation semantics. In this paper, we present PoCoFL, a policy-compliant federated learning framework that separates three aspects: (i) FL type, (ii) policy semantics, and (iii) cryptographic realisation. We provide a formalisation that captures client and aggregation requirements as policy-dependent relations. Clients prove compliance of their contributions using commitments and non-interactive zero-knowledge proofs, while aggregators prove that the recorded set of admitted contributions was processed according to the selected aggregation policy. We demonstrate PoCoFL through four formal instantiations: (i) vanilla, (ii) continual, (iii) personalised, and (iv) threshold-encrypted federated learning. We evaluate the effects of policy enforcement on the learning objectives of vanilla, personalised, and continual FL. We further implement proof-of-concept realisations of all four instantiations, demonstrating the versatility and practical feasibility of PoCoFL. Overall, these results show that PoCoFL can capture complex policy representations while remaining network-topology agnostic.

Dominik Roy George, Varesh Mishra, Aysajan Abidin

COSIC, KU Leuven, Belgium

Federated LearningPrivacy-Preserving Federated LearningFederated Learning Aggregation
AskPDF
Aug 16, 2025

FAIRVAR: Fair Federated Learning via Variance Regularization

Federated learning (FL) allows collaborative training of machine learning models across multiple parties without sharing raw data. However, heterogeneous data can cause some clients to have disproportionate influence on the global model, leading to disparities in their performance. Fairness, understood as reducing these disparities, is therefore a crucial concern in FL and has been addressed in various ways. We studied performance equitable fairness in FL, where the goal is to minimize performance disparities across clients. We evaluated several existing fairness-aware methods and introduce here a new gradient-variance-regularized method, implemented in two variants: FairGrad (approximate) and FairGrad* (exact). We theoretically characterize the connections between these methods and, empirically, on heterogeneous benchmarks, show that FairGrad and FairGrad* consistently improve fairness by reducing variance in client accuracies, while maintaining competitive or improved mean performance compared to existing fairness-aware baselines.

Zahra Kharaghani, Ali Dadras, Tommy Löfstedt

Department of Computing Science, Umeå University, Umeå, Sweden · Department of Mathematics and Mathematical Statistics, Umeå University, Umeå, Sweden

Algorithmic FairnessFederated LearningHeterogeneous Federated Learning
AskPDF