arXiv cs.AI· Sunday Oyinlola Ogundoyin, Muhammad Ikram·· 3 天前AI 评分34
多智能体 LLM 系统中拓扑驱动的信息泄露与按接收方定制的净化机制
When Agents Talk, Secrets Travel: Topology-Driven Leakage and Recipient-Conditioned Sanitization in Multi-Agent LLM Systems
AI 导读
多智能体 LLM 系统（MALS）存在传统敏感度防御无法捕捉的隐私泄露：信息经常规协作传播至未授权智能体。研究评估了 485 个场景，发现拓扑结构使向拓扑引入接收方的泄露从 0% 升至 52.2%，而表面泄露稳定在 60.6%-62.8%。提出的 RECONS 净化器将残余泄露从 96.7% 降至 29.25%，同时保留 94.1% 的授权披露。
正文 · 原文
Abstract:Multi-agent LLM systems (MALS) introduce a privacy-leakage problem that conventional sensitivity-based defenses miss, since information disclosed to one authorized agent can propagate through routine coordination to an unauthorized one, without compromise or malicious behavior. This motivates topology-aware enforcement that blocks unauthorized propagation while preserving authorized disclosure. We evaluate 485 scenarios across coordinator-worker, peer-to-peer, and hierarchical topologies, instrumenting seven communication channels and measuring surface leakage and disclosure to topology-introduced recipients. We introduce RECONS, a Detect -> Decide -> Transform sanitizer conditioning intervention on recipient and execution context rather than sensitivity alone. Topology changes where protected information propagates, with disclosure to topology-introduced recipients rising from 0% to 52.2% while surface leakage stays nearly flat at 60.6-62.8%. This effect persists across four generator models and a second orchestration framework, though magnitude and relay paths differ; framework choice separately drives a 150x leakage spread, so topology alone is not sufficient. RECONS reduces residual disclosure from 96.7% without sanitization to 29.25%, preserving 94.1% of authorized disclosures versus 20.5% under blanket redaction. Detection coverage is the dominant bottleneck, accounting for 95.5% of residual disclosures. These results show that privacy enforcement in MALS must reason about the recipient and communication path realized at runtime, not sensitivity or topology assumptions alone.
| Subjects: | Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.16039 [cs.CR] |
| (or arXiv:2610.16039v1 [cs.CR] for this version) | |
| https://doi.org/10.48550/arXiv.2610.16039 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Sunday Ogundoyin [view email]
[v1]
Mon, 12 Oct 2026 01:23:51 UTC (2,864 KB)
来源：arXiv cs.AI · arxiv.org