AI memory refers to mechanisms that allow LLMs to retain information across conversations. This includes conversation history, user preferences, and learned facts that persist beyond a single session.
Memory Types
- Session memory: Within-conversation context
- Persistent memory: Cross-session user facts
- RAG memory: Retrieved from vector databases
Memory Privacy Risks
- PII stored in memory systems
- Memory poisoning attacks
- Cross-user data leakage
- Compliance with data retention policies
Practical example
An assistant remembers a user’s preferred report format for a later session.
What to check
Test consent or authorization, retrieval scope, expiration, correction, and deletion of stored memories.
Limits and tradeoffs
Memory can preserve inaccurate or sensitive information and later reintroduce it into context.
Maintained by DriftRail. AI assisted the October 2026 content and template revision. Examples are illustrative; they are not measured product benchmarks. Suggest a correction.