OWASP Agentic
ASI06 · Memory and Context Poisoning
Persistent memory or shared context is seeded with malicious data that influences future sessions.
OWASP Top 10 for Agentic Applications (2026)
Mitigations
- Validate and expire memory entries
- Provenance tags on stored context
- Isolate memory per user/tenant
Related techniques
- LLM04 Data and Model Poisoning OWASP LLM
- LLM08 Vector and Embedding Weaknesses OWASP LLM
- AML.T0080 AI Agent Context Poisoning MITRE ATLAS
- AML.T0099 AI Agent Tool Data Poisoning MITRE ATLAS
In the news
-
DriftGuard: Online Memory-Drift Detection via Causal Attribution for LLM Agents
DriftGuard is a method for online detection of adversarial or misleading memory drift in LLM agents by estimating each retrieved memory's causal contribution.
-
Surfacing Disagreement Beats Resolving It: A Gated Ablation Study of Provenance-Weighted Memory for Multi-Agent Debate
A gated ablation study finds that preserving contradictions in shared multi-agent debate memory cuts attack success on poisoned items from 22 to 7 of 122.
Mappings are AI-assisted and checked against the framework definitions; corrections welcome.