Differential privacy (DP) is a gold-standard concept of measuring and guaranteeing privacy in data analysis. It is well-known that the cost of adding DP to deep learning model is its accuracy. However, it remains unclear how it affects robustness of the model. Standard neural networks are not robust to different input perturbations: either adversarial attacks or common corruptions. In this paper, we empirically observe an interesting trade-off between privacy and robustness of neural networks. We experimentally demonstrate that networks, trained with DP, in some settings might be even more vulnerable in comparison to non-private versions. To explore this, we extensively study different robustness measurements, including FGSM and PGD adversaries, distance to linear decision boundaries, curvature profile, and performance on a corrupted dataset. Finally, we study how the main ingredients of differentially private neural networks training, such as gradient clipping and noise addition, affect (decrease and increase) the robustness of the model.
Federated learning is emerging as a machine learning technique that trains a model across multiple decentralized parties. It is renowned for preserving privacy as the data never leaves the computational devices, and recent approaches further enhance its privacy by hiding messages transferred in encryption. However, we found that despite the efforts, federated learning remains privacy-threatening, due to its interactive nature across different parties. In this paper, we analyze the privacy threats in industrial-level federated learning frameworks with secure computation, and reveal such threats…
A Survey of Privacy Threats and Defense in Vertical Federated Learning: From Model Life Cycle Perspective
Rensselaer Polytechnic Institute（伦斯勒理工学院）；Zhejiang University（浙江大学）；AntGroup（蚂蚁集团）；Ulsan National Institute of Science and Technology（蔚山国立科学技术院）；The University of Hong Kong（香港大学）
Fraunhofer Institute for Applied and Integrated Security（弗劳恩霍夫应用与集成安全研究所）
An important problem in deep learning is the privacy and security of neural networks (NNs). Both aspects have long been considered separately. To date, it is still poorly understood how privacy enhancing training affects the robustness of NNs. This paper experimentally evaluates the impact of training with Differential Privacy (DP), a standard method for privacy preservation, on model vulnerability against a broad range of adversarial attacks. The results suggest that private models are less robust than their non-private counterparts, and that adversarial examples transfer better among DP mode…
Federated Learning (FL) is emerging as a promising paradigm of privacy-preserving machine learning, which trains an algorithm across multiple clients without exchanging their data samples. Recent works highlighted several privacy and robustness weaknesses in FL and addressed these concerns using local differential privacy (LDP) and some well-studied methods used in conventional ML, separately. However, it is still not clear how LDP affects adversarial robustness in FL. To fill this gap, this work attempts to develop a comprehensive understanding of the effects of LDP on adversarial robustness …
New Jersey Institute of Technology（新泽西理工学院）；Kent State University（肯特州立大学）；University of Oregon（俄勒冈大学）；University of Arkansas（阿肯色大学）；University of Florida（佛罗里达大学）
In this paper, we propose a novel Heterogeneous Gaussian Mechanism (HGM) to preserve differential privacy in deep neural networks, with provable robustness against adversarial examples. We first relax the constraint of the privacy budget in the traditional Gaussian Mechanism from (0, 1] to (0, \infty), with a new bound of the noise scale to preserve differential privacy. The noise in our mechanism can be arbitrarily redistributed, offering a distinctive ability to address the trade-off between model utility and privacy loss. To derive provable robustness, our HGM is applied to inject Gaussian …
WeBank（微众银行）；Hong Kong University of Science and Technology（香港科技大学）；National University of Singapore（新加坡国立大学）；Nanyang Technological University（南洋理工大学）
With the emergence of data silos and popular privacy awareness, the traditional centralized approach of training artificial intelligence (AI) models is facing strong challenges. Federated learning (FL) has recently emerged as a promising solution under this new reality. Existing FL protocol design has been shown to exhibit vulnerabilities which can be exploited by adversaries both within and without the system to compromise data privacy. It is thus of paramount importance to make FL system designers to be aware of the implications of future FL algorithm design on privacy-preservation. Currentl…
From Mean to Extreme: Formal Differential Privacy Bounds on the Success of Real-World Data Reconstruction Attacks
Institute of Machine Learning in Biomedical Imaging, Helmholtz Munich（海德堡医学院机器学习研究所）；School of Computation, Information and Technology, Technical University of Munich（慕尼黑技术大学计算、信息与技术学院）；Chair for AI in Healthcare and Medicine, Technical University of Munich (TUM) and TUM University Hospital（慕尼黑技术大学人工智能在医疗与医学中的 chair）；School of Biomedical Engineering and Imaging Sciences, King’s College London（伦敦国王学院生物医学工程与成像科学学院）；Department of Computing, Imperial College London（伦敦帝国理工学院计算机系）
Florida International University（佛罗里达国际大学）；Southern Illinois University（南伊利诺伊大学）
Federated Learning (FL) has gained widespread popularity in recent years due to the fast booming of advanced machine learning and artificial intelligence along with emerging security and privacy threats. FL enables efficient model generation from local data storage of the edge devices without revealing the sensitive data to any entities. While this paradigm partly mitigates the privacy issues of users' sensitive data, the performance of the FL process can be threatened and reached a bottleneck due to the growing cyber threats and privacy violation techniques. To expedite the proliferation of F…
Decentralized Federated Learning: A Survey on Security and Privacy
University of Windsor（温莎大学）；University of New Brunswick（新不伦瑞克大学）；University of Western Ontario（西安大略大学）；Hong Kong University of Science and Technology（香港科技大学）；WeBank（微众银行）
Andalusian Research Institute in Data Science and Computational Intelligence（安达卢西亚数据科学与计算智能研究院）；University of Granada（格拉纳达大学）
Federated learning is a machine learning paradigm that emerges as a solution to the privacy-preservation demands in artificial intelligence. As machine learning, federated learning is threatened by adversarial attacks against the integrity of the learning model and the privacy of data via a distributed approach to tackle local and global learning. This weak point is exacerbated by the inaccessibility of data in federated learning, which makes harder the protection against adversarial attacks and evidences the need to furtherance the research on defence methods to make federated learning a real…
Université Paris-Dauphine, PSL Research University（巴黎第九大学，PSL研究大学）；Institut LIST, CEA（法国原子能与替代能源委员会LIST研究所）；Université Paris-Saclay（巴黎萨克雷大学）
This short note highlights some links between two lines of research within the emerging topic of trustworthy machine learning: differential privacy and robustness to adversarial examples. By abstracting the definitions of both notions, we show that they build upon the same theoretical ground and hence results obtained so far in one domain can be transferred to the other. More precisely, our analysis is based on two key elements: probabilistic mappings (also called randomized algorithms in the differential privacy community), and the Renyi divergence which subsumes a large family of divergences…
Iowa State University（爱荷华州立大学）；University of South Florida（南佛罗里达大学）
For privacy concerns to be addressed adequately in current machine learning systems, the knowledge gap between the machine learning and privacy communities must be bridged. This article aims to provide an introduction to the intersection of both fields with special emphasis on the techniques used to protect the data.
A Critical Review on the Effectiveness and Privacy Threats of Membership Inference Attacks
Universitat Rovira i Virgili（罗维拉-维格利大学）；Department of Computer Engineering（计算机工程系）；Mathematics（数学）；CYBERCAT-Center for Cybersecurity Research of Catalonia（加泰罗尼亚网络安全研究中心）；ComSCIAM-Center for Computational Science and Applied Mathematics（计算科学与应用数学研究中心）
University of Technology Sydney（悉尼科技大学）；Macquarie University（麦考瑞大学）；University of South Wales（南威尔士大学）；Federation University（联邦大学）
Security and privacy of the users have become significant concerns due to the involvement of the Internet of things (IoT) devices in numerous applications. Cyber threats are growing at an explosive pace making the existing security and privacy measures inadequate. Hence, everyone on the Internet is a product for hackers. Consequently, Machine Learning (ML) algorithms are used to produce accurate outputs from large complex databases, where the generated outputs can be used to predict and detect vulnerabilities in IoT-based systems. Furthermore, Blockchain (BC) techniques are becoming popular in…
Hong Kong University of Science and Technology（香港科技大学）；Southern University of Science and Technology（南方科技大学）；WeBank Co. Ltd（微众银行）
Matrix Factorization has been very successful in practical recommendation applications and e-commerce. Due to data shortage and stringent regulations, it can be hard to collect sufficient data to build performant recommender systems for a single company. Federated learning provides the possibility to bridge the data silos and build machine learning models without compromising privacy and security. Participants sharing common users or items collaboratively build a model over data from all the participants. There have been some works exploring the application of federated learning to recommender…
Every year at NeurIPS, machine learning researchers gather and discuss exciting applications of machine learning in areas such as public health, disaster response, climate change, education, and more. However, many of these same researchers are expressing growing concern about applications of machine learning for surveillance (Nanayakkara et al., 2021). This paper presents a brief overview of strategies for resisting these surveillance technologies and calls for greater collaboration between machine learning and human-computer interaction researchers to address the threats that these technolog…
Cybersecurity information sharing (CIS) is envisioned to protect organizations more effectively from advanced cyber attacks. However, a completely automated CIS platform is not widely adopted. The major challenges are: (1) the absence of a robust cyber threat language (CTL) and (2) the concerns over data privacy. This work introduces Cybersecurity Information Exchangewith Privacy (CYBEX-P), as a CIS framework, to tackle these challenges. CYBEX-P allows organizations to share heterogeneous data with granular, attribute based privacy control. It correlates the data to automatically generate intu…
In recent years, researchers have extensively studied adversarial robustness in a variety of threat models, including L_0, L_1, L_2, and L_infinity-norm bounded adversarial attacks. However, attacks bounded by fractional L_p "norms" (quasi-norms defined by the L_p distance with 0<p<1) have yet to be thoroughly considered. We proactively propose a defense with several desirable properties: it provides provable (certified) robustness, scales to ImageNet, and yields deterministic (rather than high-probability) certified guarantees when applied to quantized data (e.g., images). Our technique for f…
A major drawback of adversarially robust models, in particular for large scale datasets like ImageNet, is the extremely long training time compared to standard ones. Moreover, models should be robust not only to one $l_p$-threat model but ideally to all of them. In this paper we propose Extreme norm Adversarial Training (E-AT) for multiple-norm robustness which is based on geometric properties of $l_p$-balls. E-AT costs up to three times less than other adversarial training methods for multiple-norm robustness. Using E-AT we show that for ImageNet a single epoch and for CIFAR-10 three epochs a…
Existing defenses against adversarial examples such as adversarial training typically assume that the adversary will conform to a specific or known threat model, such as $\ell_p$ perturbations within a fixed budget. In this paper, we focus on the scenario where there is a mismatch in the threat model assumed by the defense during training, and the actual capabilities of the adversary at test time. We ask the question: if the learner trains against a specific "source" threat model, when can we expect robustness to generalize to a stronger unknown "target" threat model during test-time? Our key …
A key challenge in adversarial robustness is the lack of a precise mathematical characterization of human perception, used in the very definition of adversarial attacks that are imperceptible to human eyes. Most current attacks and defenses try to avoid this issue by considering restrictive adversarial threat models such as those bounded by $L_2$ or $L_\infty$ distance, spatial perturbations, etc. However, models that are robust against any of these restrictive threat models are still fragile against other threat models. To resolve this issue, we propose adversarial training against the set of…
