Memory as an Attack Surface in LLM Agents: A Study on Multiple-Choice Question Answering

Abstract

AI agents extend conventional large language model (LLM) applications by integrating language understanding with task execution, external tool use, and memory mechanisms. While memory allows agents to retain prior interactions and provide more personalized and context-aware responses, it also introduces a new vulnerability: information stored in memory can influence future outputs even when the current query is clean. In this paper, we investigate memory manipulation in LLM-based agents for mult

Categories

Framework mappings

Suggested from the entry's categories.

@misc{sakib2026memory,
  title = {{Memory as an Attack Surface in LLM Agents: A Study on Multiple-Choice Question Answering}},
  author = {Shahnewaz Karim Sakib and Anindya Bijoy Das},
  year = {2026},
  month = jun,
  eprint = {2606.29030},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2606.29030}
}