MétaCan
Menu
Back to cohort
Record W4411270140 · doi:10.2139/ssrn.5288768

Machine Unlearning Doesn't Do What You Think: Lessons for Generative AI Policy and Research

2025· preprint· en· W4411270140 on OpenAlexaff
A. Feder Cooper, Christopher A. Choquette-Choo, Miranda Bogen, Kevin Klyman, Matthew Jagielski, Katja Filippova, Ken Liu, Alexandra Chouldechova, Jamie Hayes, Yangsibo Huang, Niloofar Mireshghallah, Ilia Shumailov, Eleni Triantafillou, Peter Kairouz, Nicole Mitchell, Percy Liang, Daniel E. Ho, Yejin Choi, Sanmi Koyejo, Fernando Delgado, James Grimmelmann, Vitaly Shmatikov, Christopher De, Solon Barocas, Amy Cyphert, Mark A. Lemley, Pamela Samuelson, danah boyd, Jennifer Wortman Vaughan, Miles Brundage, David Bau, Seth Neel, Abigail Jacobs, Andreas Terzis, Hanna Wallach, Nicolas Papernot, Katherine Lee

Bibliographic record

VenueSSRN Electronic Journal · 2025
Typepreprint
Languageen
FieldSocial Sciences
TopicEconomic and Technological Developments in Russia
Canadian institutionsUniversity of Toronto
Fundersnot available
KeywordsGenerative grammarComputer scienceArtificial intelligence

Abstract

fetched live from OpenAlex

"Machine unlearning" is a popular proposed solution for mitigating the existence of content in an AI model that is problematic for legal or moral reasons, including privacy, copyright, safety, and more. For example, unlearning is often invoked as a solution for removing the effects of targeted information from a generative-AI model's parameters, e.g., a particular individual's personal data or the inclusion of copyrighted content in the model's training data. Unlearning is also proposed as a way to prevent a model from generating targeted types of information in its outputs, e.g., generations that closely resemble a particular individual's data or reflect the concept of "Spiderman." Both of these goals—the targeted removal of information from a model and the targeted suppression of information from a model's outputs—present various technical and substantive challenges. We provide a framework for ML researchers and policymakers to think rigorously about these challenges, identifying several mismatches between the goals of unlearning and feasible implementations. These mismatches explain why unlearning is not a general-purpose solution for circumscribing generative-AI model behavior in service of broader positive impact.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.015
metaresearch head score (Gemma)0.052
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesMetaresearch
Consensus categoriesnone
DomainCandidate signal: Methods · Consensus signal: none
Study designCandidate signal: Theoretical or conceptual · Consensus signal: Theoretical or conceptual
GenreCandidate signal: Commentary · Consensus signal: Commentary
Teacher disagreement score0.985
Threshold uncertainty score0.079

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0150.052
Meta-epidemiology (narrow)0.0010.000
Meta-epidemiology (broad)0.0020.001
Bibliometrics0.0010.001
Science and technology studies0.0020.025
Scholarly communication0.0070.015
Open science0.0020.003
Research integrity0.0060.008
Insufficient payload (model declined to judge)0.0070.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.053
GPT teacher head0.419
Teacher spread0.365 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

Study designTheoretical or conceptual
DomainMethods
GenreCommentary

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations1
Published2025
Admission routes1
Has abstractno

Explore more

Same venueSSRN Electronic JournalSame topicEconomic and Technological Developments in RussiaFrench-language works237,207