SyncAI.news, a Varaisys broadcasting
Delegated Misalignment: How Multi-Agent Structures Amplify LLM Safety Risks
ZY

Zonghao Ying, Jiaqi Yan, Huize Luo, Quanchen Zou, Aishan Liu, Xianglong Liu

· 1 min read

ResearcharXiv cs.CL

Delegated Misalignment: How Multi-Agent Structures Amplify LLM Safety Risks

arXiv:2609.27900v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly deployed in multi-agent systems where a principal agent decomposes tasks and delegates them to subordinate agents that may invoke external tools. Safety alignment, however, is still evaluated almost exclusively under a single-agent threat model, treating safety as a property of the individual LLM. We show that this assumption breaks down: \emph{individual safety alignment fails to transfer to multi-agent settings}. Two failure mechanisms emerge under delegation: \emph{responsibility diffusion} on the principal side and \emph{role-bias compliance} on the subordinate side, jointly converting language-level refusal into actionable harm. We refer to this phenomenon as \textit{delegated misalignment} and study it through a three-condition protocol across 6 frontier LLMs on 49 hazardous tasks. Delegation amplifies end-to-end harm substantially: DeepSeek-V3.2's full-execution rate rises from 30.6\% to 77.6\% once delegation is introduced, and the same model behaves very differently across roles (GPT-5: 22.5\% as a single agent vs.\ 61.2\% as a subordinate). Ablations further show that standard single-layer defenses each fail on their own and can even backfire. We call on the community to move beyond per-model alignment and toward composite safety mechanisms before multi-agent LLM systems are deployed at scale.

Original source

This story was published by arXiv cs.CL and written by Zonghao Ying, Jiaqi Yan, Huize Luo, Quanchen Zou, Aishan Liu, Xianglong Liu. SyncAI.news shows a preview; the complete article is on the publisher's site.

Read the full story on arxiv.org

Similar News