arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.05112cs.CRcs.AI
收藏

从善意中释放恶意:良性数据对机器遗忘的威胁

Releasing Malevolence from Benevolence: The Menace of Benign Data on Machine Unlearning

Zhejiang University(浙江大学);Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security(杭州高新区(滨江)区块链与数据安全研究院);CSIRO’s Data61(澳大利亚联邦科学与工业研究组织Data61);King Abdullah University of Science and Technology(阿卜杜拉国王科技大学);Shanghai Jiao Tong University(上海交通大学)

Binhao Ma, Tianhang Zheng, Hongsheng Hu, Di Wang, Shuo Wang, Zhongjie Ba, Zhan Qin, Kui Ren

AI总结:

本文提出 Unlearning Usability Attack,利用不超过训练集 1% 且可通过投毒检测的良性数据发起遗忘请求,最高使模型准确率下降 50%,揭示机器遗忘的新型安全威胁。

摘要:

在大量真实或合成数据上训练的机器学习模型,往往能在各个领域取得出色的预测性能。然而,这种效用伴随着日益增长的隐私担忧,因为训练数据可能包含敏感信息。为解决这些担忧,机器遗忘被提出用于从模型中擦除特定数据样本。尽管一些遗忘技术能够以低成本高效移除数据,但近期研究强调了相关漏洞:恶意用户可能请求对被操纵的数据进行遗忘,从而破坏模型。尽管这些攻击有效,但扰动数据与原始训练数据不同,无法通过哈希验证。现有的机器遗忘攻击还存在实际局限,需要大量额外知识和资源。为弥补当前遗忘攻击的空白,我们提出了 Unlearning Usability Attack(遗忘可用性攻击)。这种与模型无关、与遗忘方法无关且预算友好的攻击,将数据分布信息蒸馏到一小部分良性数据中。由于这些数据对模型训练具有积极影响,自动投毒检测工具会将其识别为良性。虽然它们对机器学习而言是良性的,但遗忘这些数据会显著削弱模型信息。我们的评估表明,遗忘这些不超过总训练数据 1% 的良性数据,可使模型准确率降低最高达 50%。此外,我们的研究结果表明,精心准备的良性数据对近期遗忘技术构成挑战,因为擦除这些合成实例比常规数据需要更高资源。这些见解强调,未来研究有必要在机器遗忘背景下重新思考“数据投毒”。

英文摘要:

Machine learning models trained on vast amounts of real or synthetic data often achieve outstanding predictive performance across various domains. However, this utility comes with increasing concerns about privacy, as the training data may include sensitive information. To address these concerns, machine unlearning has been proposed to erase specific data samples from models. While some unlearning techniques efficiently remove data at low costs, recent research highlights vulnerabilities where malicious users could request unlearning on manipulated data to compromise the model. Despite these attacks' effectiveness, perturbed data differs from original training data, failing hash verification. Existing attacks on machine unlearning also suffer from practical limitations and require substantial additional knowledge and resources. To fill the gaps in current unlearning attacks, we introduce the Unlearning Usability Attack. This model-agnostic, unlearning-agnostic, and budget-friendly attack distills data distribution information into a small set of benign data. These data are identified as benign by automatic poisoning detection tools due to their positive impact on model training. While benign for machine learning, unlearning these data significantly degrades model information. Our evaluation demonstrates that unlearning this benign data, comprising no more than 1% of the total training data, can reduce model accuracy by up to 50%. Furthermore, our findings show that well-prepared benign data poses challenges for recent unlearning techniques, as erasing these synthetic instances demands higher resources than regular data. These insights underscore the need for future research to reconsider "data poisoning" in the context of machine unlearning.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑