When Agents Talk, Secrets Travel: Topology-Driven Leakage and Recipient-Conditioned Sanitization in Multi-Agent LLM Systems
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.
ProofPaper ↗
Key points
- 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%.
Sources (1)
- [1]When Agents Talk, Secrets Travel: Topology-Driven Leakage and Recipient-Conditioned Sanitization in Multi-Agent LLM SystemsarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 12, 01:23 AM
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.
Extractive summary: sentences quoted from the sources.