Search

Though there is much interest in fair AI systems, the problem of fairness noncompliance -- which concerns whether fair models are used in practice -- has received lesser attention. Zero-Knowledge Proofs of Fairness (ZKPoF) address fairness noncompliance by allowing a service provider to verify to external parties that their model serves diverse demographics equitably, with guaranteed confidentiality over proprietary model parameters and data. They have great potential for building public trust and effective AI regulation, but no previous techniques for ZKPoF are fit for real-world deployment. We present OATH, the first ZKPoF framework that is (i) deployably efficient with client-facing communication comparable to in-the-clear ML as a Service query answering, and an offline audit phase that verifies an asymptotically constant quantity of answered queries, (ii) deployably flexible with modularity for any score-based classifier given a zero-knowledge proof of correct inference, (iii) deployably secure with an end-to-end security model that guarantees confidentiality and fairness across training, inference, and audits. We show that OATH obtains strong robustness against malicious adversaries at concretely efficient parameter settings. Notably, OATH provides a 1343x improvement to runtime over previous work for neural network ZKPoF, and scales up to much larger models -- even DNNs with tens of millions of parameters.

We present Nebula, a system for differential private histogram estimation of data distributed among clients. Nebula enables clients to locally subsample and encode their data such that an untrusted server learns only data values that meet an aggregation threshold to satisfy differential privacy guarantees. Compared with other private histogram estimation systems, Nebula uniquely achieves all of the following: \textit{i)} a strict upper bound on privacy leakage; \textit{ii)} client privacy under realistic trust assumptions; \textit{iii)} significantly better utility compared to standard local differential privacy systems; and \textit{iv)} avoiding trusted third-parties, multi-party computation, or trusted hardware. We provide both a formal evaluation of Nebula's privacy, utility and efficiency guarantees, along with an empirical evaluation on three real-world datasets. We demonstrate that clients can encode and upload their data efficiently (only 0.0058 seconds running time and 0.0027 MB data communication) and privately (strong differential privacy guarantees $\varepsilon=1$). On the United States Census dataset, the Nebula's untrusted aggregation server estimates histograms with above 88\% better utility than the existing local deployment of differential privacy. Additionally, we describe a variant that allows clients to submit multi-dimensional data, with similar privacy, utility, and performance. Finally, we provide an open source implementation of Nebula.

Prior Membership Inference Attacks (MIAs) on pre-trained Large Language Models (LLMs), adapted from classification model attacks, fail due to ignoring the generative process of LLMs across token sequences. In this paper, we present a novel attack that adapts MIA statistical tests to the perplexity dynamics of subsequences within a data point. Our method significantly outperforms prior loss-based approaches, revealing context-dependent memorization patterns in pre-trained LLMs.

Large Language Models (LLMs) have shown greatly enhanced performance in recent years, attributed to increased size and extensive training data. This advancement has led to widespread interest and adoption across industries and the public. However, training data memorization in Machine Learning models scales with model size, particularly concerning for LLMs. Memorized text sequences have the potential to be directly leaked from LLMs, posing a serious threat to data privacy. Various techniques have been developed to attack LLMs and extract their training data. As these models continue to grow, this issue becomes increasingly critical. To help researchers and policymakers understand the state of knowledge around privacy attacks and mitigations, including where more work is needed, we present the first SoK on data privacy for LLMs. We (i) identify a taxonomy of salient dimensions where attacks differ on LLMs, (ii) systematize existing attacks, using our taxonomy of dimensions to highlight key trends, (iii) survey existing mitigation strategies, highlighting their strengths and limitations, and (iv) identify key gaps, demonstrating open problems and areas for concern., Comment: 15 pages

Federated learning (FL) is a framework for users to jointly train a machine learning model. FL is promoted as a privacy-enhancing technology (PET) that provides data minimization: data never "leaves" personal devices and users share only model updates with a server (e.g., a company) coordinating the distributed training. While prior work showed that in vanilla FL a malicious server can extract users' private data from the model updates, in this work we take it further and demonstrate that a malicious server can reconstruct user data even in hardened versions of the protocol. More precisely, we propose an attack against FL protected with distributed differential privacy (DDP) and secure aggregation (SA). Our attack method is based on the introduction of sybil devices that deviate from the protocol to expose individual users' data for reconstruction by the server. The underlying root cause for the vulnerability to our attack is a power imbalance: the server orchestrates the whole protocol and users are given little guarantees about the selection of other users participating in the protocol. Moving forward, we discuss requirements for privacy guarantees in FL. We conclude that users should only participate in the protocol when they trust the server or they apply local primitives such as local DP, shifting power away from the server. Yet, the latter approaches come at significant overhead in terms of performance degradation of the trained model, making them less likely to be deployed in practice.

Private multi-winner voting is the task of revealing $k$-hot binary vectors satisfying a bounded differential privacy (DP) guarantee. This task has been understudied in machine learning literature despite its prevalence in many domains such as healthcare. We propose three new DP multi-winner mechanisms: Binary, $\tau$, and Powerset voting. Binary voting operates independently per label through composition. $\tau$ voting bounds votes optimally in their $\ell_2$ norm for tight data-independent guarantees. Powerset voting operates over the entire binary vector by viewing the possible outcomes as a power set. Our theoretical and empirical analysis shows that Binary voting can be a competitive mechanism on many tasks unless there are strong correlations between labels, in which case Powerset voting outperforms it. We use our mechanisms to enable privacy-preserving multi-label learning in the central setting by extending the canonical single-label technique: PATE. We find that our techniques outperform current state-of-the-art approaches on large, real-world healthcare data and standard multi-label benchmarks. We further enable multi-label confidential and private collaborative (CaPC) learning and show that model performance can be significantly improved in the multi-site setting., Comment: Accepted at PoPETS 2023

Adversarial attacks modify images with perturbations that change the prediction of classifiers. These modified images, known as adversarial examples, expose the vulnerabilities of deep neural network classifiers. In this paper, we investigate the predictability of the mapping between the classes predicted for original images and for their corresponding adversarial examples. This predictability relates to the possibility of retrieving the original predictions and hence reversing the induced misclassification. We refer to this property as the reversibility of an adversarial attack, and quantify reversibility as the accuracy in retrieving the original class or the true class of an adversarial example. We present an approach that reverses the effect of an adversarial attack on a classifier using a prior set of classification results. We analyse the reversibility of state-of-the-art adversarial attacks on benchmark classifiers and discuss the factors that affect the reversibility.

In this paper, we study the problem of learning Graph Neural Networks (GNNs) with Differential Privacy (DP). We propose a novel differentially private GNN based on Aggregation Perturbation (GAP), which adds stochastic noise to the GNN's aggregation function to statistically obfuscate the presence of a single edge (edge-level privacy) or a single node and all its adjacent edges (node-level privacy). Tailored to the specifics of private learning, GAP's new architecture is composed of three separate modules: (i) the encoder module, where we learn private node embeddings without relying on the edge information; (ii) the aggregation module, where we compute noisy aggregated node embeddings based on the graph structure; and (iii) the classification module, where we train a neural network on the private aggregations for node classification without further querying the graph edges. GAP's major advantage over previous approaches is that it can benefit from multi-hop neighborhood aggregations, and guarantees both edge-level and node-level DP not only for training, but also at inference with no additional costs beyond the training's privacy budget. We analyze GAP's formal privacy guarantees using R\'enyi DP and conduct empirical experiments over three real-world graph datasets. We demonstrate that GAP offers significantly better accuracy-privacy trade-offs than state-of-the-art DP-GNN approaches and naive MLP-based baselines. Our code is publicly available at https://github.com/sisaman/GAP., Comment: Accepted at USENIX Security '23

Sharing real-world speech utterances is key to the training and deployment of voice-based services. However, it also raises privacy risks as speech contains a wealth of personal data. Speaker anonymization aims to remove speaker information from a speech utterance while leaving its linguistic and prosodic attributes intact. State-of-the-art techniques operate by disentangling the speaker information (represented via a speaker embedding) from these attributes and re-synthesizing speech based on the speaker embedding of another speaker. Prior research in the privacy community has shown that anonymization often provides brittle privacy protection, even less so any provable guarantee. In this work, we show that disentanglement is indeed not perfect: linguistic and prosodic attributes still contain speaker information. We remove speaker information from these attributes by introducing differentially private feature extractors based on an autoencoder and an automatic speech recognizer, respectively, trained using noise layers. We plug these extractors in the state-of-the-art anonymization pipeline and generate, for the first time, private speech utterances with a provable upper bound on the speaker information they contain. We evaluate empirically the privacy and utility resulting from our differentially private speaker anonymization approach on the LibriSpeech data set. Experimental results show that the generated utterances retain very high utility for automatic speech recognition training and inference, while being much better protected against strong adversaries who leverage the full knowledge of the anonymization process to try to infer the speaker identity.

Recent years have seen a surge in the popularity of acoustics-enabled personal devices powered by machine learning. Yet, machine learning has proven to be vulnerable to adversarial examples. A large number of modern systems protect themselves against such attacks by targeting artificiality, i.e., they deploy mechanisms to detect the lack of human involvement in generating the adversarial examples. However, these defenses implicitly assume that humans are incapable of producing meaningful and targeted adversarial examples. In this paper, we show that this base assumption is wrong. In particular, we demonstrate that for tasks like speaker identification, a human is capable of producing analog adversarial examples directly with little cost and supervision: by simply speaking through a tube, an adversary reliably impersonates other speakers in eyes of ML models for speaker identification. Our findings extend to a range of other acoustic-biometric tasks such as liveness detection, bringing into question their use in security-critical settings in real life, such as phone banking., Comment: Published at USENIX Security 2023 https://www.usenix.org/conference/usenixsecurity23/presentation/ahmed

In federated learning (FL), data does not leave personal devices when they are jointly training a machine learning model. Instead, these devices share gradients, parameters, or other model updates, with a central party (e.g., a company) coordinating the training. Because data never "leaves" personal devices, FL is often presented as privacy-preserving. Yet, recently it was shown that this protection is but a thin facade, as even a passive, honest-but-curious attacker observing gradients can reconstruct data of individual users contributing to the protocol. In this work, we show a novel data reconstruction attack which allows an active and dishonest central party to efficiently extract user data from the received gradients. While prior work on data reconstruction in FL relies on solving computationally expensive optimization problems or on making easily detectable modifications to the shared model's architecture or parameters, in our attack the central party makes inconspicuous changes to the shared model's weights before sending them out to the users. We call the modified weights of our attack trap weights. Our active attacker is able to recover user data perfectly, i.e., with zero error, even when this data stems from the same class. Recovery comes with near-zero costs: the attack requires no complex optimization objectives. Instead, our attacker exploits inherent data leakage from model gradients and simply amplifies this effect by maliciously altering the weights of the shared model through the trap weights. These specificities enable our attack to scale to fully-connected and convolutional deep neural networks trained with large mini-batches of data. For example, for the high-dimensional vision dataset ImageNet, we perfectly reconstruct more than 50% of the training data points from mini-batches as large as 100 data points.

Speaker identification models are vulnerable to carefully designed adversarial perturbations of their input signals that induce misclassification. In this work, we propose a white-box steganography-inspired adversarial attack that generates imperceptible adversarial perturbations against a speaker identification model. Our approach, FoolHD, uses a Gated Convolutional Autoencoder that operates in the DCT domain and is trained with a multi-objective loss function, in order to generate and conceal the adversarial perturbation within the original audio files. In addition to hindering speaker identification performance, this multi-objective loss accounts for human perception through a frame-wise cosine similarity between MFCC feature vectors extracted from the original and adversarial audio files. We validate the effectiveness of FoolHD with a 250-speaker identification x-vector network, trained using VoxCeleb, in terms of accuracy, success rate, and imperceptibility. Our results show that FoolHD generates highly imperceptible adversarial audio files (average PESQ scores above 4.30), while achieving a success rate of 99.6% and 99.2% in misleading the speaker identification model, for untargeted and targeted settings, respectively., Comment: https://fsepteixeira.github.io/FoolHD/

We present an adversarial framework to craft perturbations that mislead classifiers by accounting for the image content and the semantics of the labels. The proposed framework combines a structure loss and a semantic adversarial loss in a multi-task objective function to train a fully convolutional neural network. The structure loss helps generate perturbations whose type and magnitude are defined by a target image processing filter. The semantic adversarial loss considers groups of (semantic) labels to craft perturbations that prevent the filtered image {from} being classified with a label in the same group. We validate our framework with three different target filters, namely detail enhancement, log transformation and gamma correction filters; and evaluate the adversarially filtered images against three classifiers, ResNet50, ResNet18 and AlexNet, pre-trained on ImageNet. We show that the proposed framework generates filtered images with a high success rate, robustness, and transferability to unseen classifiers. We also discuss objective and subjective evaluations of the adversarial perturbations., Comment: 13 pages

Adversarial perturbations can be added to images to protect their content from unwanted inferences. These perturbations may, however, be ineffective against classifiers that were not {seen} during the generation of the perturbation, or against defenses {based on re-quantization, median filtering or JPEG compression. To address these limitations, we present an adversarial attack {that is} specifically designed to protect visual content against { unseen} classifiers and known defenses. We craft perturbations using an iterative process that is based on the Fast Gradient Signed Method and {that} randomly selects a classifier and a defense, at each iteration}. This randomization prevents an undesirable overfitting to a specific classifier or defense. We validate the proposed attack in both targeted and untargeted settings on the private classes of the Places365-Standard dataset. Using ResNet18, ResNet50, AlexNet and DenseNet161 {as classifiers}, the performance of the proposed attack exceeds that of eleven state-of-the-art attacks. The implementation is available at https://github.com/smartcameras/RP-FGSM/.

We present DarkneTZ, a framework that uses an edge device's Trusted Execution Environment (TEE) in conjunction with model partitioning to limit the attack surface against Deep Neural Networks (DNNs). Increasingly, edge devices (smartphones and consumer IoT devices) are equipped with pre-trained DNNs for a variety of applications. This trend comes with privacy risks as models can leak information about their training data through effective membership inference attacks (MIAs). We evaluate the performance of DarkneTZ, including CPU execution time, memory usage, and accurate power consumption, using two small and six large image classification models. Due to the limited memory of the edge device's TEE, we partition model layers into more sensitive layers (to be executed inside the device TEE), and a set of layers to be executed in the untrusted part of the operating system. Our results show that even if a single layer is hidden, we can provide reliable model privacy and defend against state of the art MIAs, with only 3% performance overhead. When fully utilizing the TEE, DarkneTZ provides model protections with up to 10% overhead., Comment: 13 pages, 8 figures, accepted to ACM MobiSys 2020

Machine Learning as a Service (MLaaS) operators provide model training and prediction on the cloud. MLaaS applications often rely on centralised collection and aggregation of user data, which could lead to significant privacy concerns when dealing with sensitive personal data. To address this problem, we propose PrivEdge, a technique for privacy-preserving MLaaS that safeguards the privacy of users who provide their data for training, as well as users who use the prediction service. With PrivEdge, each user independently uses their private data to locally train a one-class reconstructive adversarial network that succinctly represents their training data. As sending the model parameters to the service provider in the clear would reveal private information, PrivEdge secret-shares the parameters among two non-colluding MLaaS providers, to then provide cryptographically private prediction services through secure multi-party computation techniques. We quantify the benefits of PrivEdge and compare its performance with state-of-the-art centralised architectures on three privacy-sensitive image-based tasks: individual identification, writer identification, and handwritten letter recognition. Experimental results show that PrivEdge has high precision and recall in preserving privacy, as well as in distinguishing between private and non-private images. Moreover, we show the robustness of PrivEdge to image compression and biased training data. The source code is available at https://github.com/smartcameras/PrivEdge., Comment: IEEE Transactions on Information Forensics and Security (TIFS)

Adversarial attacks that generate small L_p-norm perturbations to mislead classifiers have limited success in black-box settings and with unseen classifiers. These attacks are also not robust to defenses that use denoising filters and to adversarial training procedures. Instead, adversarial attacks that generate unrestricted perturbations are more robust to defenses, are generally more successful in black-box settings and are more transferable to unseen classifiers. However, unrestricted perturbations may be noticeable to humans. In this paper, we propose a content-based black-box adversarial attack that generates unrestricted perturbations by exploiting image semantics to selectively modify colors within chosen ranges that are perceived as natural by humans. We show that the proposed approach, ColorFool, outperforms in terms of success rate, robustness to defense frameworks and transferability, five state-of-the-art adversarial attacks on two different tasks, scene and object classification, when attacking three state-of-the-art deep neural networks using three standard datasets. The source code is available at https://github.com/smartcameras/ColorFool., Comment: Conference on Computer Vision and Pattern Recognition (CVPR2020)

Adversarial examples are intentionally perturbed images that mislead classifiers. These images can, however, be easily detected using denoising algorithms, when high-frequency spatial perturbations are used, or can be noticed by humans, when perturbations are large. In this paper, we propose EdgeFool, an adversarial image enhancement filter that learns structure-aware adversarial perturbations. EdgeFool generates adversarial images with perturbations that enhance image details via training a fully convolutional neural network end-to-end with a multi-task loss function. This loss function accounts for both image detail enhancement and class misleading objectives. We evaluate EdgeFool on three classifiers (ResNet-50, ResNet-18 and AlexNet) using two datasets (ImageNet and Private-Places365) and compare it with six adversarial methods (DeepFool, SparseFool, Carlini-Wagner, SemanticAdv, Non-targeted and Private Fast Gradient Sign Methods). Code is available at https://github.com/smartcameras/EdgeFool.git.

Pre-trained Deep Neural Network (DNN) models are increasingly used in smartphones and other user devices to enable prediction services, leading to potential disclosures of (sensitive) information from training data captured inside these models. Based on the concept of generalization error, we propose a framework to measure the amount of sensitive information memorized in each layer of a DNN. Our results show that, when considered individually, the last layers encode a larger amount of information from the training data compared to the first layers. We find that, while the neuron of convolutional layers can expose more (sensitive) information than that of fully connected layers, the same DNN architecture trained with different datasets has similar exposure per layer. We evaluate an architecture to protect the most sensitive layers within the memory limits of Trusted Execution Environment (TEE) against potential white-box membership inference attacks without the significant computational overhead., Comment: 5 pages, 6 figures, CCS PPML workshop

Recently, there has been a wealth of effort devoted to the design of secure protocols for machine learning tasks. Much of this is aimed at enabling secure prediction from highly-accurate Deep Neural Networks (DNNs). However, as DNNs are trained on data, a key question is how such models can be also trained securely. The few prior works on secure DNN training have focused either on designing custom protocols for existing training algorithms, or on developing tailored training algorithms and then applying generic secure protocols. In this work, we investigate the advantages of designing training algorithms alongside a novel secure protocol, incorporating optimizations on both fronts. We present QUOTIENT, a new method for discretized training of DNNs, along with a customized secure two-party protocol for it. QUOTIENT incorporates key components of state-of-the-art DNN training such as layer normalization and adaptive gradient methods, and improves upon the state-of-the-art in DNN training in two-party computation. Compared to prior work, we obtain an improvement of 50X in WAN time and 6% in absolute accuracy.

We propose a cloud-based filter trained to block third parties from uploading privacy-sensitive images of others to online social media. The proposed filter uses Distributed One-Class Learning, which decomposes the cloud-based filter into multiple one-class classifiers. Each one-class classifier captures the properties of a class of privacy-sensitive images with an autoencoder. The multi-class filter is then reconstructed by combining the parameters of the one-class autoencoders. The training takes place on edge devices (e.g. smartphones) and therefore users do not need to upload their private and/or sensitive images to the cloud. A major advantage of the proposed filter over existing distributed learning approaches is that users cannot access, even indirectly, the parameters of other users. Moreover, the filter can cope with the imbalanced and complex distribution of the image content and the independent probability of addition of new users. We evaluate the performance of the proposed distributed filter using the exemplar task of blocking a user from sharing privacy-sensitive images of other users. In particular, we validate the behavior of the proposed multi-class filter with non-privacy-sensitive images, the accuracy when the number of classes increases, and the robustness to attacks when an adversary user has access to privacy-sensitive images of other users.

We present and evaluate Deep Private-Feature Extractor (DPFE), a deep model which is trained and evaluated based on information theoretic constraints. Using the selective exchange of information between a user's device and a service provider, DPFE enables the user to prevent certain sensitive information from being shared with a service provider, while allowing them to extract approved information using their model. We introduce and utilize the log-rank privacy, a novel measure to assess the effectiveness of DPFE in removing sensitive information and compare different models based on their accuracy-privacy tradeoff. We then implement and evaluate the performance of DPFE on smartphones to understand its complexity, resource demands, and efficiency tradeoffs. Our results on benchmark image datasets demonstrate that under moderate resource utilization, DPFE can achieve high accuracy for primary tasks while preserving the privacy of sensitive features.

Deep neural networks are increasingly being used in a variety of machine learning applications applied to rich user data on the cloud. However, this approach introduces a number of privacy and efficiency challenges, as the cloud operator can perform secondary inferences on the available data. Recently, advances in edge processing have paved the way for more efficient, and private, data processing at the source for simple tasks and lighter models, though they remain a challenge for larger, and more complicated models. In this paper, we present a hybrid approach for breaking down large, complex deep models for cooperative, privacy-preserving analytics. We do this by breaking down the popular deep architectures and fine-tune them in a particular way. We then evaluate the privacy benefits of this approach based on the information exposed to the cloud service. We also asses the local inference cost of different layers on a modern handset for mobile applications. Our evaluations show that by using certain kind of fine-tuning and embedding techniques and at a small processing costs, we can greatly reduce the level of information available to unintended tasks applied to the data feature on the cloud, and hence achieving the desired tradeoff between privacy and performance., Comment: arXiv admin note: substantial text overlap with arXiv:1703.02952

Internet of Things (IoT) devices and applications are being deployed in our homes and workplaces. These devices often rely on continuous data collection to feed machine learning models. However, this approach introduces several privacy and efficiency challenges, as the service operator can perform unwanted inferences on the available data. Recently, advances in edge processing have paved the way for more efficient, and private, data processing at the source for simple tasks and lighter models, though they remain a challenge for larger, and more complicated models. In this paper, we present a hybrid approach for breaking down large, complex deep neural networks for cooperative, privacy-preserving analytics. To this end, instead of performing the whole operation on the cloud, we let an IoT device to run the initial layers of the neural network, and then send the output to the cloud to feed the remaining layers and produce the final result. In order to ensure that the user's device contains no extra information except what is necessary for the main task and preventing any secondary inference on the data, we introduce Siamese fine-tuning. We evaluate the privacy benefits of this approach based on the information exposed to the cloud service. We also assess the local inference cost of different layers on a modern handset. Our evaluations show that by using Siamese fine-tuning and at a small processing cost, we can greatly reduce the level of unnecessary, potentially sensitive information in the personal data, and thus achieving the desired trade-off between utility, privacy, and performance., Comment: To appear in IEEE Internet of Things Journal