Skip to content
Search
paperJune 2026Unreviewed

From Untrusted Input to Trusted Memory: A Systematic Study of Memory Poisoning Attacks in LLM Agents

Pritam Dash, Tongyu Ge, Aditi Jain, Tanmay Shah, Zhiwei Shang

View resourcePDF

Abstract

Memory is a core component of AI agents, enabling them to accumulate knowledge across interactions and improve performance. However, persistent memory introduces the risk of memory poisoning, where a single adversarial memory write can exert long-term influence over agent behavior. We present a systematic study of memory poisoning in LLM-based agents. We identify four memory write channels and nine structural vulnerabilities in model capabilities, system prompt design, and agent system architect

Categories

Data PoisoningAgentic ThreatsMemory Security

Framework mappings

OWASP Top 10 for LLM Applications
  • LLM04Data and Model Poisoning
OWASP Top 10 for Agentic Applications
  • ASI06Memory & Context Poisoning
MITRE ATLAS
  • AML.T0020Poison Training Data
  • AML.T0080AI Agent Context Poisoning

Suggested from the entry's categories.

Cite

@misc{dash2026from,
  title = {{From Untrusted Input to Trusted Memory: A Systematic Study of Memory Poisoning Attacks in LLM Agents}},
  author = {Pritam Dash and Tongyu Ge and Aditi Jain and Tanmay Shah and Zhiwei Shang},
  year = {2026},
  month = jun,
  eprint = {2606.04329},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2606.04329}
}