面向安全多智能体系统的私有知识检索与私有推理研究
On Private Knowledge Retrieval and Private Inference for secure multi-agent systems
AI 导读
一篇 arXiv 论文针对 LLM 多智能体系统中的语义记忆投毒攻击，提出私有知识检索与私有推理机制，通过在外部知识库引入签名系统实现安全设计。该架构中智能体内置类 Prolog 的推理系统（用 Horn 子句表示知识），私有访问外部签名知识库，并给出基于 k-匿名性与信息论隐私（IT-PIR）的多种 Python 实现。
正文 · 原文
Abstract:Multi-agent systems provide a convenient framework for the massive deployment of agents based on Large Language Models (LLMs). It provides a sound framework where agents can interact with existing software to perform actions in an autonomous way. Agentic AI is the new term to denote these AI-based autonomous systems.
In real multi-agent systems, agents need to deal with malicious agents some of which can exploit security vulnerabilities to do harm. A memory poisoning attack is an example of exploiting a vulnerability to modify agents knowledge and, thus, affect its behavior. In an agent, there are different types of memory systems (semantic, episodic, and short-term) which have their own characteristics and for which different memory poisoning attacks may be developed.
In this paper we focus on semantic memory, we discuss its attack, and mitigation strategies. We introduce private knowledge retrieval and private inference. We propose a signature system in the knowledge base to prevent from undesired manipulation. This should avoid memory poisoning attacks by introducing security by design in the construction of the semantic memory. In our architecture, an agent contains a local Prolog-like inference system (i.e., using Horn clauses to represent knowledge) which accesses an external knowledge base to get the required knowledge. This externally signed knowledge base is accessed privately. We provide different implementations in Python using different privacy levels including k-anonymity and information-theoretical privacy (IT-PIR).
| Subjects: | Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) |
| ACM classes: | I.2.11 |
| Cite as: | arXiv:2611.13696 [cs.CR] |
| (or arXiv:2611.13696v1 [cs.CR] for this version) | |
| https://doi.org/10.48550/arXiv.2611.13696 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Vicenc Torra [view email]
[v1]
Tue, 10 Nov 2026 07:43:02 UTC (19 KB)
来源：arXiv cs.AI · arxiv.org