面向多智能体人工智能系统的治理感知联邦学习：可信协作式人工智能的策略驱动框架

arXiv cs.AI 论文
federated-learning multi-agent-systems governance trustworthy-ai fairness auditability

摘要

提出了治理感知联邦学习框架GAFL，为多智能体AI系统在协作训练中加入证据验证、公平性评估、策略执行与防篡改审计日志，并在跨孤岛原型上验证了其对伪造证据与恶意更新的识别能力。

arXiv:2610.38411v1 Announce Type: new 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.
查看原文
查看缓存全文

缓存时间: 2026/10/31 09:15 

# Governance-Aware Federated Learning for Multi-Agent AI Systems: A Policy-Driven Framework for Trustworthy Collaborative AI
Source: [https://arxiv.org/abs/2610.38411](https://arxiv.org/abs/2610.38411)
[View PDF](https://arxiv.org/pdf/2610.38411)

> 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\.

## Submission history

From: Kiranmayee Janardhan Dr \[[view email](https://arxiv.org/show-email/4e7f15b0/2610.38411)\] **\[v1\]**Wed, 28 Oct 2026 07:57:10 UTC \(610 KB\)

相似文章

多智能体 LLM 谈判中的宪法规门控与确定性恢复：针对有状态对抗性 Gatekeeper 的消融实验

arXiv cs.CL

该论文研究多智能体 LLM 谈判系统中的三部分控制栈：5 条支柱运行时宪章、4 层 swarm 架构（Director、多数投票、Monitor、schema 硬门控）以及 Cognitive Annealing（确定性死锁检测与原子上下文清除恢复），并针对一个有状态对抗性 Gatekeeper 进行消融实验，发现确定性恢复机制在死锁场景中显著优于纯 LLM 引导。

利用可验证信息奖励训练大语言模型进行多智能体协作

Hugging Face Daily Papers

这篇名为CoSIR的预印本提出通过可验证信息奖励（基于贝叶斯惊异度降低）对LLM进行自博弈强化学习后训练，使模型显式学习多智能体协作能力，并在7个未见过的协作基准上取得最优表现，同时基本保持原有单任务性能。