Evidence · DiggingBeagle record
Proxy unlearning uses representational overlap for unrequested knowledge erasure
Paper detailing an attacker who selects proxy deletion requests to cause unintended collateral model knowledge loss during unlearning.
- Published
- Nov 11, 2026
- Source role
- primary disclosure
Evidence record
Paper detailing an attacker who selects proxy deletion requests to cause unintended collateral model knowledge loss during unlearning.
Claim-level citations (2)
- supportsA permissible model-unlearning request could erase other knowledge that the requester had no authority to delete: The proxy-unlearning attack exploits representational overlap. An attacker who can request deletion of one permitted group of entities selects proxy facts whose hidden representations overlap a different protected target, so a removal procedure trained to forget the proxies also degrades retention of the target. In controlled model tests, selected proxies caused larger degradation of protected factual associations than randomly chosen deletion targets; a superposition-aware protection method attempts to screen overlapping requests and restore unrelated facts.
Proxy Unlearning paper, threat model, proxy selection, Section VII controlled comparisons
- supportsA permissible model-unlearning request could erase other knowledge that the requester had no authority to delete: The attack presupposes access to a model-unlearning request workflow and a susceptible update method. It does not demonstrate exploitation of any deployed commercial forgetting service or justify refusing legitimate privacy deletion requests indiscriminately. The failure class is authorization of effect: approving which data to forget is distinct from proving only that data was changed in the actual model state.
Primary disclosure: results, scope and limitations