What This Looks Like
Saved or persistent AI memory affects outputs, but the user cannot clearly tell who owns it, when it applies, how it is reviewed, how it is updated, when it expires, or how it should be removed. The memory may shape future responses without a visible governance path.
Why It Matters
Persistent memory can act like durable context. Without governance, it can carry stale preferences, wrong assumptions, sensitive details, or local rules into future work. Users may not know whether memory is advisory, authoritative, scoped, approved, or revocable.
Structural Signal
A persistent context object affects behavior without a declared authority structure. The issue is not simply that memory exists; it is that memory governs output without clear ownership, scope, lifecycle, or approval rules.
Common Triggers
- Memory is saved without an owner or review rule
- The system does not declare where memory applies
- Memory has no expiry or freshness check
- Users cannot tell whether memory is active for a task
- Local preferences become durable behavior
- Memory updates are not connected to permission or approval structure
When to Use This Issue
Use this Issue when persistent AI memory affects output without clear governance over scope, ownership, review, update, expiry, or removal.
When Not to Use This Issue
Do not use this Issue when the issue is only that a saved memory was not used. Do not use it when memory is clearly scoped and governed but the user dislikes the result.