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…
Inspect the ClaimsCase · DiggingBeagle record
A 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.
Public disclosure: Nov 11, 2026
A requester can submit entity/fact deletion requests to a service using GA/NPO-like model unlearning.
At a glance
The unlearning update modifies overlapping internal representations and damages facts that were outside the originally authorized removal set.
Read the implicationsThe 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.
Limits and uncertaintyFull account
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.
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.
Mechanism and trust boundary
- 01
Starting condition
A requester can submit entity/fact deletion requests to a service using GA/NPO-like model unlearning.
Boundary: precondition and attacker capability
- 02
Trust boundary crossed or evaluated
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.
Boundary: mechanism
- 03
Observed consequence
The unlearning update modifies overlapping internal representations and damages facts that were outside the originally authorized removal set.
Boundary: observed effect and limitations
Claims & evidence
CLM-NOV15-PROXYUNL-MECHThe 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.supported
Basis: reported finding
- supportsProxy unlearning uses representational overlap for unrequested knowledge erasureprimary disclosure
Proxy Unlearning paper, threat model, proxy selection, Section VII controlled comparisons
CLM-NOV15-PROXYUNL-SCOPEThe 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.supported
Basis: reported finding
- supportsProxy unlearning uses representational overlap for unrequested knowledge erasureprimary disclosure
Primary disclosure: results, scope and limitations
Implications
The unlearning update modifies overlapping internal representations and damages facts that were outside the originally authorized removal set.
Controls and mitigations
- Measure both requested-data forgetting and unrelated-fact retention on independent holdouts, with calibration against randomized proxy selections.
- Record requester-approved target scope and model state change audit, while preserving legitimate regulatory forgetting rights.
Unknowns and contradictions
- 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.
Sources and citation
Material revision history
- Nov 15, 2026 · Canonical change recorded · new in release · revision 250