Machine Unlearning Doesn't Do What You Think: Lessons for Generative AI Policy, Research, and Practice
Cooper, A. Feder, Choquette-Choo, Christopher A., Bogen, Miranda, Jagielski, Matthew, Filippova, Katja, Liu, Ken Ziyu, Chouldechova, Alexandra, Hayes, Jamie, Huang, Yangsibo, Mireshghallah, Niloofar, Shumailov, Ilia, Triantafillou, Eleni, Kairouz, Peter, Mitchell, Nicole, Liang, Percy, Ho, Daniel E., Choi, Yejin, Koyejo, Sanmi, Delgado, Fernando, Grimmelmann, James, Shmatikov, Vitaly, De Sa, Christopher, Barocas, Solon, Cyphert, Amy, Lemley, Mark, boyd, danah, Vaughan, Jennifer Wortman, Brundage, Miles, Bau, David, Neel, Seth, Jacobs, Abigail Z., Terzis, Andreas, Wallach, Hanna, Papernot, Nicolas, Lee, Katherine
–arXiv.org Artificial Intelligence
We articulate fundamental mismatches between technical methods for machine unlearning in Generative AI, and documented aspirations for broader impact that these methods could have for law and policy. These aspirations are both numerous and varied, motivated by issues that pertain to 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 in-copyright expression of Spiderman that was included 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 thinking rigorously about these challenges, which enables us to be clear about why unlearning is not a general-purpose solution for circumscribing generative-AI model behavior in service of broader positive impact. We aim for conceptual clarity and to encourage more thoughtful communication among machine learning (ML), law, and policy experts who seek to develop and apply technical methods for compliance with policy objectives.
arXiv.org Artificial Intelligence
Dec-9-2024
- Country:
- Europe (1.00)
- North America > United States
- California (0.28)
- Genre:
- Research Report > Promising Solution (0.48)
- Industry:
- Education > Curriculum
- Subject-Specific Education (1.00)
- Government > Regional Government
- Europe Government (0.93)
- North America Government > United States Government (1.00)
- Health & Medicine > Therapeutic Area (0.93)
- Information Technology > Security & Privacy (1.00)
- Law
- Law Enforcement & Public Safety > Crime Prevention & Enforcement (0.92)
- Education > Curriculum
- Technology: