# Securing LLM-Agent Long-Term Memory Against Poisoning: Non-Malleable, Origin-Bound Authority with Machine-Checked Guarantees

arXiv: https://arxiv.org/abs/2606.24322 · HTML: https://mlsanity.com/paper/2606.24322/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- Placed in Agent threats, guardrails and governance, one of 16 directions in LLM agents and tool use.
- Builds on 33 of its 42 references that are in the map.
- First posted on arXiv 23 Jun 2026.

## Placed in

- LLM agents and tool use / [Agent threats, guardrails and governance](https://mlsanity.com/agents/agent-threats-guardrails-and-governance/)

## Builds on (most cited first, 30 of 33)

- [Generative Agents: Interactive Simulacra of Human Behavior](https://mlsanity.com/paper/2304.03442/) (2023)
- [MemGPT: Towards LLMs as Operating Systems](https://mlsanity.com/paper/2310.08560/) (2023)
- [Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory](https://mlsanity.com/paper/2504.19413/) (2025)
- [Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training](https://mlsanity.com/paper/2401.05566/) (2024)
- [AgentPoison: Red-teaming LLM Agents via Poisoning Memory or Knowledge Bases](https://mlsanity.com/paper/2407.12784/) (2024)
- [Defeating Prompt Injections by Design](https://mlsanity.com/paper/2503.18813/) (2025)
- [PoisonedRAG: Knowledge Corruption Attacks to Retrieval-Augmented Generation of Large Language Models](https://mlsanity.com/paper/2402.07867/) (2024)
- [Memory Injection Attacks on LLM Agents via Query-Only Interaction](https://mlsanity.com/paper/2503.03704/) (2025)
- [Watch Out for Your Agents! Investigating Backdoor Threats to LLM-Based Agents](https://mlsanity.com/paper/2402.11208/) (2024)
- [Securing AI Agents with Information-Flow Control](https://mlsanity.com/paper/2505.23643/) (2025)
- [MemoryGraft: Persistent Compromise of LLM Agents via Poisoned Experience Retrieval](https://mlsanity.com/paper/2512.16962/) (2025)
- [Taming Various Privilege Escalation in LLM-Based Agent Systems: A Mandatory Access Control Framework](https://mlsanity.com/paper/2601.11893/) (2026)
- [Memory Poisoning Attack and Defense on Memory Based LLM-Agents](https://mlsanity.com/paper/2601.05504/) (2026)
- [TrustRAG: Enhancing Robustness and Trustworthiness in Retrieval-Augmented Generation](https://mlsanity.com/paper/2501.00879/) (2025)
- [A Survey on Long-Term Memory Security in LLM Agents: Attacks, Defenses, and Governance Across the Memory Lifecycle](https://mlsanity.com/paper/2604.16548/) (2026)
- [Hidden in Memory: Sleeper Memory Poisoning in LLM Agents](https://mlsanity.com/paper/2605.15338/) (2026)
- [From Storage to Steering: Memory Control Flow Attacks on LLM Agents](https://mlsanity.com/paper/2603.15125/) (2026)
- [MemMachine: A Ground-Truth-Preserving Memory System for Personalized AI Agents](https://mlsanity.com/paper/2604.04853/) (2026)
- [MemLineage: Lineage-Guided Enforcement for LLM Agent Memory](https://mlsanity.com/paper/2605.14421/) (2026)
- [Ghost in the Agent: Redefining Information Flow Tracking for LLM Agents](https://mlsanity.com/paper/2604.23374/) (2026)
- [DrunkAgent: Stealthy Memory Corruption in LLM-Powered Recommender Agents](https://mlsanity.com/paper/2503.23804/) (2025)
- [Trojan Hippo: Weaponizing Agent Memory for Data Exfiltration](https://mlsanity.com/paper/2605.01970/) (2026)
- [A Systematic Survey of Security Threats and Defenses in LLM-Based AI Agents: A Layered Attack Surface Framework](https://mlsanity.com/paper/2604.23338/) (2026)
- [Improving Google A2A Protocol: Protecting Sensitive Data and Mitigating Unintended Harms in Multi-Agent Systems](https://mlsanity.com/paper/2505.12490/) (2025)
- [Security Analysis of Agentic AI Communication Protocols: A Comparative Evaluation](https://mlsanity.com/paper/2511.03841/) (2025)
- [MemMorph: Tool Hijacking in LLM Agents via Memory Poisoning](https://mlsanity.com/paper/2605.26154/) (2026)
- [SuperLocalMemory: Privacy-Preserving Multi-Agent Memory with Bayesian Trust Defense Against Memory Poisoning](https://mlsanity.com/paper/2603.02240/) (2026)
- [Visual Inception: Compromising Long-term Planning in Agentic Recommenders via Multimodal Memory Poisoning](https://mlsanity.com/paper/2604.16966/) (2026)
- [Nonmalleable Information Flow: Technical Report](https://mlsanity.com/paper/1708.08596/) (2017)
- [Hijacking Agent Memory: Stealthy Trojan Attacks Through Conversational Interaction](https://mlsanity.com/paper/2605.29960/) (2026)

## Built on by (most cited first, 18 of 18)

- [Autonomous Research Agents: A Survey of AI Scientists and the Verification Gap](https://mlsanity.com/paper/2608.05179/) (2026)
- [SkillJack: Persistent Skill Backdoors in Self-Evolving Agents](https://mlsanity.com/paper/2608.03509/) (2026)
- [MemSecBench: Tracking Agent Memory Poisoning from Persistence to Consequence and Repair](https://mlsanity.com/paper/2607.27080/) (2026)
- [Ground Truth First: A Longitudinal Evaluation Instrument for Agent Memory, and the Tenure Crossover in Memory-Architecture Rankings](https://mlsanity.com/paper/2607.21962/) (2026)
- [When Stale Constraints Go Unchecked: Budgeted Verification Failures in Inherited Agent Memory](https://mlsanity.com/paper/2608.25553/) (2026)
- [Epistemic Sybil Resistance: Multiplying AI Agents Without Multiplying Evidence](https://mlsanity.com/paper/2609.01873/) (2026)
- [Protocol-Level Attacks on Agentic Commerce Platforms: A Cross-Platform Taxonomy, AIP-Bench, and Unified Defense](https://mlsanity.com/paper/2607.21824/) (2026)
- [Utility Under Attack: Agent Memory Poisoning and the Limits of Content Screening and Provenance Ranking](https://mlsanity.com/paper/2608.21230/) (2026)
- [Revoked but Still Authoritative: An Empirical Study of Revocation Enforcement in Agent-Memory Systems](https://mlsanity.com/paper/2609.08258/) (2026)
- [Silence Is Endorsement: Verification-Status Laundering in LLM Agent Pipelines](https://mlsanity.com/paper/2609.20211/) (2026)
- [Activity Frames: Deterministic Screen-Activity Compilation for Agent Memory and Replay](https://mlsanity.com/paper/2608.05784/) (2026)
- [Transferable End-to-End Optimization for Indirect Long-Term Memory Poisoning in LLM Agents](https://mlsanity.com/paper/2609.00523/) (2026)
- [Evolutionary Safety of Recursive Self-Improving AI: Taxonomy, Risk Discovery, and Evaluation](https://mlsanity.com/paper/2609.31186/) (2026)
- [RoutePrism: Tracing Construction Order Effects in Agent Memory](https://mlsanity.com/paper/2609.34160/) (2026)
- [Audience-Bound Persistent Memory: Authorization Across the Memory Lifecycle](https://mlsanity.com/paper/2609.36373/) (2026)
- [The Cognitive Continuity Test: Verifying Governed State Transitions in Persistent AI Agents](https://mlsanity.com/paper/2610.00132/) (2026)
- [When Evidence Changes: Evaluating Memory Repair and Re-reading in Language-Model Agents](https://mlsanity.com/paper/2610.03902/) (2026)
- [Self-Propagating Misalignment in LLM Agents, and Why Auditing or Disabling Memory Is Not Enough](https://mlsanity.com/paper/2610.04083/) (2026)

## Abstract (arXiv)

LLM agents increasingly rely on persistent long-term memory, which creates a critical vulnerability that we study here: memory poisoning. An adversary can store untrusted content in one session that later steers a consequential action, such as a payment, a setting change, or data exfiltration, in a future session. Existing defenses base a memory item's authority to act on either its content (detection or trust-scoring) or its derivation history (lineage). We show that both signals are malleable. An attacker can launder an untrusted origin through three channels specific to LLM agents: the agent's own summarization, a trusted-tool echo, and manufactured corroboration. Each makes the content look benign and breaks or flips its derivation edge to ``trusted.'' We formalize malleability for the memory write-retrieve-act pipeline and prove a machine-checked separation theorem. No content- or lineage-based defense is sound under laundering (T1), write-time origin binding is necessary (T2), and non-malleable origin-bound authority with Sybil-resistant corroboration-gated elevation is sufficient (T3). Our construction, TMA-NM (Tamper-evident Memory Authority, Non-Malleable), instantiates non-malleable information-flow control (IFC) for LLM-agent memory. A cross-defense, cross-attack, and cross-model benchmark over eight frontier models shows that existing defenses fail exactly where the theory predicts (up to 68% laundering attack-success), while TMA-NM reaches 0% attack success on both direct and laundering attacks across all models and channels, at full legitimate utility. We release the benchmark, harness, and machine-checked TLA+ models to support reproducibility.