Supporting Trustworthy AI Through Machine Unlearning

Author:

Hine EmmieORCID,Novelli ClaudioORCID,Taddeo MariarosariaORCID,Floridi LucianoORCID

Abstract

AbstractMachine unlearning (MU) is often analyzed in terms of how it can facilitate the “right to be forgotten.” In this commentary, we show that MU can support the OECD’s five principles for trustworthy AI, which are influencing AI development and regulation worldwide. This makes it a promising tool to translate AI principles into practice. We also argue that the implementation of MU is not without ethical risks. To address these concerns and amplify the positive impact of MU, we offer policy recommendations across six categories to encourage the research and uptake of this potentially highly influential new technology.

Funder

Alma Mater Studiorum - Università di Bologna

Publisher

Springer Science and Business Media LLC

Reference60 articles.

1. Achille, A., Kearns, M., Klingenberg, C., & Soatto, S. (2023). AI model disgorgement: Methods and choices. Proceedings of the National Academy of Sciences., 121(18), e2307304121.

2. Albergotti, R. (2023). The secret history of Elon Musk, Sam Altman, and OpenAI. Semafor. March 24, 2023. https://www.semafor.com/article/03/24/2023/the-secret-history-of-elon-musk-sam-altman-and-openai

3. Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021). On the dangers of stochastic parrots: Can language models be too big? In Proceedings of the 2021 (ACM conference on fairness, accountability, and transparency (FAccT) (pp. 610-623).

4. Blistein, J. (2023). Sarah Silverman leads class action copyright suit against ChatGPT. Rolling Stone (blog). https://www.rollingstone.com/culture/culture-news/sarah-silverman-copoyright-suit-chatgpt-open-ai-1234785472/

5. Bourtoule, L., Chandrasekaran, V., Choquette-Choo, C. A., Jia, H., Travers, A., Zhang, B., & Papernot, N. (2021). Machine unlearning. In 2021 IEEE symposium on security and privacy (S&P) (pp. 141-159). IEEE.

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3