Skip to main content

2 docs tagged with "embedding inversion"

View all tags

Multi-tenant RAG retrieval leakage: vectors aren't anonymous, and filtering by user isn't always isolation

In one sentence: in multi-tenant RAG, two intuitions both fail — "vectorized means anonymized" and "filtered by user means isolated." In some studied embedding settings a vector can be inverted back to approximate original text (including PII), so don't treat it as anonymization; retrieval ranks by similarity, not permission, so if the ACL runs after retrieval instead of before it, the most relevant private chunk may come from another tenant; and a single boundary bug in a shared cache / long-term memory can carry one user's private context into another's session.