Recognises an erasure stopping at production
Erasure requests change the system of record and leave copies, with no rule for them.
Can AI do this?
Not settled yet The benchmark tested this, but the models' answers do not settle yet whether they master it. 0 of 4 models in the current benchmark master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 85% | Not shown |
| gemini-3.5-flash-lite | Low cost | 82% | Not shown |
| gpt-6-luna | Low cost | 85% | Not shown |
| claude-haiku-5-5 | Low cost | 85% | Not shown |
Mastered means a probability of at least 90% that the model has this knobit, estimated from its graded answers. “Not shown” means the answers do not settle it either way. How the models are tested.
What it looks like in practice
Expected result: The defect named on the erasure leaving copies untouched, and its fix or the question for its owner
How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: Erasure procedure covers DS-004 only; warehouse, extracts and backups not mentioned Acceptable: Every copy covered, or residual copies recorded with their expiry
Also known as: Erasure that stops at production
Related knobits
Builds on
All 14 knobits of Implement retention, archiving and purge.
Source: Review checklist 04-retention-archiving, row 6 (GDPR Art. 17)