Jinyuan Jia

Jinyuan Jia, Wenjie Qu, Neil Zhenqiang Gong

Multi-label classification, which predicts a set of labels for an input, has many applications. However, multiple recent studies showed that multi-label classification is vulnerable to adversarial examples. In particular, an attacker can manipulate the labels predicted by a multi-label classifier for an input via adding carefully crafted, human-imperceptible perturbation to it. Existing provable defenses for multi-class classification achieve sub-optimal provable robustness guarantees when generalized to multi-label classification. In this work, we propose MultiGuard, the first provably robust defense against adversarial examples to multi-label classification. Our MultiGuard leverages randomized smoothing, which is the state-of-the-art technique to build provably robust classifiers. Specifically, given an arbitrary multi-label classifier, our MultiGuard builds a smoothed multi-label classifier via adding random noise to the input. We consider isotropic Gaussian noise in this work. Our major theoretical contribution is that we show a certain number of ground truth labels of an input are provably in the set of labels predicted by our MultiGuard when the $\ell_2$-norm of the adversarial perturbation added to the input is bounded. Moreover, we design an algorithm to compute our provable robustness guarantees. Empirically, we evaluate our MultiGuard on VOC 2007, MS-COCO, and NUS-WIDE benchmark datasets. Our code is available at: \url{https://github.com/quwenjie/MultiGuard}

Fengqing Jiang, Zhangchen Xu, Luyao Niu et al. · uw

Large language models (LLMs) are increasingly deployed as the service backend for LLM-integrated applications such as code completion and AI-powered search. LLM-integrated applications serve as middleware to refine users' queries with domain-specific knowledge to better inform LLMs and enhance the responses. Despite numerous opportunities and benefits, LLM-integrated applications also introduce new attack surfaces. Understanding, minimizing, and eliminating these emerging attack surfaces is a new area of research. In this work, we consider a setup where the user and LLM interact via an LLM-integrated application in the middle. We focus on the communication rounds that begin with user's queries and end with LLM-integrated application returning responses to the queries, powered by LLMs at the service backend. For this query-response protocol, we identify potential vulnerabilities that can originate from the malicious application developer or from an outsider threat initiator that is able to control the database access, manipulate and poison data that are high-risk for the user. Successful exploits of the identified vulnerabilities result in the users receiving responses tailored to the intent of a threat initiator. We assess such threats against LLM-integrated applications empowered by OpenAI GPT-3.5 and GPT-4. Our empirical results show that the threats can effectively bypass the restrictions and moderation policies of OpenAI, resulting in users receiving responses that contain bias, toxic content, privacy risk, and disinformation. To mitigate those threats, we identify and define four key properties, namely integrity, source identification, attack detectability, and utility preservation, that need to be satisfied by a safe LLM-integrated application. Based on these properties, we develop a lightweight, threat-agnostic defense that mitigates both insider and outsider threats.

Xiaoyu Cao, Zaixi Zhang, Jinyuan Jia et al.

Due to its distributed nature, federated learning is vulnerable to poisoning attacks, in which malicious clients poison the training process via manipulating their local training data and/or local model updates sent to the cloud server, such that the poisoned global model misclassifies many indiscriminate test inputs or attacker-chosen ones. Existing defenses mainly leverage Byzantine-robust federated learning methods or detect malicious clients. However, these defenses do not have provable security guarantees against poisoning attacks and may be vulnerable to more advanced attacks. In this work, we aim to bridge the gap by proposing FLCert, an ensemble federated learning framework, that is provably secure against poisoning attacks with a bounded number of malicious clients. Our key idea is to divide the clients into groups, learn a global model for each group of clients using any existing federated learning method, and take a majority vote among the global models to classify a test input. Specifically, we consider two methods to group the clients and propose two variants of FLCert correspondingly, i.e., FLCert-P that randomly samples clients in each group, and FLCert-D that divides clients to disjoint groups deterministically. Our extensive experiments on multiple datasets show that the label predicted by our FLCert for a test input is provably unaffected by a bounded number of malicious clients, no matter what poisoning attacks they use.

Hongbin Liu, Jinyuan Jia, Neil Zhenqiang Gong

Contrastive learning pre-trains an image encoder using a large amount of unlabeled data such that the image encoder can be used as a general-purpose feature extractor for various downstream tasks. In this work, we propose PoisonedEncoder, a data poisoning attack to contrastive learning. In particular, an attacker injects carefully crafted poisoning inputs into the unlabeled pre-training data, such that the downstream classifiers built based on the poisoned encoder for multiple target downstream tasks simultaneously classify attacker-chosen, arbitrary clean inputs as attacker-chosen, arbitrary classes. We formulate our data poisoning attack as a bilevel optimization problem, whose solution is the set of poisoning inputs; and we propose a contrastive-learning-tailored method to approximately solve it. Our evaluation on multiple datasets shows that PoisonedEncoder achieves high attack success rates while maintaining the testing accuracy of the downstream classifiers built upon the poisoned encoder for non-attacker-chosen inputs. We also evaluate five defenses against PoisonedEncoder, including one pre-processing, three in-processing, and one post-processing defenses. Our results show that these defenses can decrease the attack success rate of PoisonedEncoder, but they also sacrifice the utility of the encoder or require a large clean pre-training dataset.

Jinghuai Zhang, Hongbin Liu, Jinyuan Jia et al.

Contrastive learning (CL) pre-trains general-purpose encoders using an unlabeled pre-training dataset, which consists of images or image-text pairs. CL is vulnerable to data poisoning based backdoor attacks (DPBAs), in which an attacker injects poisoned inputs into the pre-training dataset so the encoder is backdoored. However, existing DPBAs achieve limited effectiveness. In this work, we take the first step to analyze the limitations of existing backdoor attacks and propose new DPBAs called CorruptEncoder to CL. CorruptEncoder introduces a new attack strategy to create poisoned inputs and uses a theory-guided method to maximize attack effectiveness. Our experiments show that CorruptEncoder substantially outperforms existing DPBAs. In particular, CorruptEncoder is the first DPBA that achieves more than 90% attack success rates with only a few (3) reference images and a small poisoning ratio 0.5%. Moreover, we also propose a defense, called localized cropping, to defend against DPBAs. Our results show that our defense can reduce the effectiveness of DPBAs, but it sacrifices the utility of the encoder, highlighting the need for new defenses.

Jinghuai Zhang, Jinyuan Jia, Hongbin Liu et al.

Point cloud classification is an essential component in many security-critical applications such as autonomous driving and augmented reality. However, point cloud classifiers are vulnerable to adversarially perturbed point clouds. Existing certified defenses against adversarial point clouds suffer from a key limitation: their certified robustness guarantees are probabilistic, i.e., they produce an incorrect certified robustness guarantee with some probability. In this work, we propose a general framework, namely PointCert, that can transform an arbitrary point cloud classifier to be certifiably robust against adversarial point clouds with deterministic guarantees. PointCert certifiably predicts the same label for a point cloud when the number of arbitrarily added, deleted, and/or modified points is less than a threshold. Moreover, we propose multiple methods to optimize the certified robustness guarantees of PointCert in three application scenarios. We systematically evaluate PointCert on ModelNet and ScanObjectNN benchmark datasets. Our results show that PointCert substantially outperforms state-of-the-art certified defenses even though their robustness guarantees are probabilistic.

Hongbin Liu, Wenjie Qu, Jinyuan Jia et al.

Classifiers in supervised learning have various security and privacy issues, e.g., 1) data poisoning attacks, backdoor attacks, and adversarial examples on the security side as well as 2) inference attacks and the right to be forgotten for the training data on the privacy side. Various secure and privacy-preserving supervised learning algorithms with formal guarantees have been proposed to address these issues. However, they suffer from various limitations such as accuracy loss, small certified security guarantees, and/or inefficiency. Self-supervised learning is an emerging technique to pre-train encoders using unlabeled data. Given a pre-trained encoder as a feature extractor, supervised learning can train a simple yet accurate classifier using a small amount of labeled training data. In this work, we perform the first systematic, principled measurement study to understand whether and when a pre-trained encoder can address the limitations of secure or privacy-preserving supervised learning algorithms. Our key findings are that a pre-trained encoder substantially improves 1) both accuracy under no attacks and certified security guarantees against data poisoning and backdoor attacks of state-of-the-art secure learning algorithms (i.e., bagging and KNN), 2) certified security guarantees of randomized smoothing against adversarial examples without sacrificing its accuracy under no attacks, 3) accuracy of differentially private classifiers, and 4) accuracy and/or efficiency of exact machine unlearning.

Runpeng Geng, Yanting Wang, Chenlong Yin et al.

Long context LLMs are vulnerable to prompt injection, where an attacker can inject an instruction in a long context to induce an LLM to generate an attacker-desired output. Existing prompt injection defenses are designed for short contexts. When extended to long-context scenarios, they have limited effectiveness. The reason is that an injected instruction constitutes only a very small portion of a long context, making the defense very challenging. In this work, we propose PISanitizer, which first pinpoints and sanitizes potential injected tokens (if any) in a context before letting a backend LLM generate a response, thereby eliminating the influence of the injected instruction. To sanitize injected tokens, PISanitizer builds on two observations: (1) prompt injection attacks essentially craft an instruction that compels an LLM to follow it, and (2) LLMs intrinsically leverage the attention mechanism to focus on crucial input tokens for output generation. Guided by these two observations, we first intentionally let an LLM follow arbitrary instructions in a context and then sanitize tokens receiving high attention that drive the instruction-following behavior of the LLM. By design, PISanitizer presents a dilemma for an attacker: the more effectively an injected instruction compels an LLM to follow it, the more likely it is to be sanitized by PISanitizer. Our extensive evaluation shows that PISanitizer can successfully prevent prompt injection, maintain utility, outperform existing defenses, is efficient, and is robust to optimization-based and strong adaptive attacks. The code is available at https://github.com/sleeepeer/PISanitizer.

Yuqi Jia, Zedian Shao, Yupei Liu et al.

Large Language Models (LLMs) are vulnerable to prompt injection attacks, and several defenses have recently been proposed, often claiming to mitigate these attacks successfully. However, we argue that existing studies lack a principled approach to evaluating these defenses. In this paper, we argue the need to assess defenses across two critical dimensions: (1) effectiveness, measured against both existing and adaptive prompt injection attacks involving diverse target and injected prompts, and (2) general-purpose utility, ensuring that the defense does not compromise the foundational capabilities of the LLM. Our critical evaluation reveals that prior studies have not followed such a comprehensive evaluation methodology. When assessed using this principled approach, we show that existing defenses are not as successful as previously reported. This work provides a foundation for evaluating future defenses and guiding their development. Our code and data are available at: https://github.com/PIEval123/PIEval.

Yanting Wang, Wei Zou, Runpeng Geng et al.

Long context large language models (LLMs) are deployed in many real-world applications such as RAG, agent, and broad LLM-integrated applications. Given an instruction and a long context (e.g., documents, PDF files, webpages), a long context LLM can generate an output grounded in the provided context, aiming to provide more accurate, up-to-date, and verifiable outputs while reducing hallucinations and unsupported claims. This raises a research question: how to pinpoint the texts (e.g., sentences, passages, or paragraphs) in the context that contribute most to or are responsible for the generated output by an LLM? This process, which we call context traceback, has various real-world applications, such as 1) debugging LLM-based systems, 2) conducting post-attack forensic analysis for attacks (e.g., prompt injection attack, knowledge corruption attacks) to an LLM, and 3) highlighting knowledge sources to enhance the trust of users towards outputs generated by LLMs. When applied to context traceback for long context LLMs, existing feature attribution methods such as Shapley have sub-optimal performance and/or incur a large computational cost. In this work, we develop TracLLM, the first generic context traceback framework tailored to long context LLMs. Our framework can improve the effectiveness and efficiency of existing feature attribution methods. To improve the efficiency, we develop an informed search based algorithm in TracLLM. We also develop contribution score ensemble/denoising techniques to improve the accuracy of TracLLM. Our evaluation results show TracLLM can effectively identify texts in a long context that lead to the output of an LLM. Our code and data are at: https://github.com/Wang-Yanting/TracLLM.

Dong Liang, Yuhao Liu, Jinyuan Jia et al.

Being able to edit panoramic images is crucial for creating realistic 360° visual experiences. However, existing perspective-based image editing methods fail to model the spatial structure of panoramas. Conventional cube-map decompositions attempt to overcome this problem but inevitably break global consistency due to their mismatch with spherical geometry. Motivated by this insight, we reformulate panoramic editing directly in the equirectangular projection (ERP) domain and present World-Shaper, a unified geometry-aware framework that bridges panoramic generation and editing within a single editing-centric design. To overcome the scarcity of paired data, we adopt a generate-then-edit paradigm, where controllable panoramic generation serves as an auxiliary stage to synthesize diverse paired examples for supervised editing learning. To address geometric distortion, we introduce a geometry-aware learning strategy that explicitly enforces position-aware shape supervision and implicitly internalizes panoramic priors through progressive training. Extensive experiments on our new benchmark, PEBench, demonstrate that our method achieves superior geometric consistency, editing fidelity, and text controllability compared to SOTA methods, enabling coherent and flexible 360° visual world creation with unified editing control. Code, model, and data will be released at our project page: https://world-shaper-project.github.io/

Jinyuan Jia, Yupei Liu, Neil Zhenqiang Gong

Self-supervised learning in computer vision aims to pre-train an image encoder using a large amount of unlabeled images or (image, text) pairs. The pre-trained image encoder can then be used as a feature extractor to build downstream classifiers for many downstream tasks with a small amount of or no labeled training data. In this work, we propose BadEncoder, the first backdoor attack to self-supervised learning. In particular, our BadEncoder injects backdoors into a pre-trained image encoder such that the downstream classifiers built based on the backdoored image encoder for different downstream tasks simultaneously inherit the backdoor behavior. We formulate our BadEncoder as an optimization problem and we propose a gradient descent based method to solve it, which produces a backdoored image encoder from a clean one. Our extensive empirical evaluation results on multiple datasets show that our BadEncoder achieves high attack success rates while preserving the accuracy of the downstream classifiers. We also show the effectiveness of BadEncoder using two publicly available, real-world image encoders, i.e., Google's image encoder pre-trained on ImageNet and OpenAI's Contrastive Language-Image Pre-training (CLIP) image encoder pre-trained on 400 million (image, text) pairs collected from the Internet. Moreover, we consider defenses including Neural Cleanse and MNTD (empirical defenses) as well as PatchGuard (a provable defense). Our results show that these defenses are insufficient to defend against BadEncoder, highlighting the needs for new defenses against our BadEncoder. Our code is publicly available at: https://github.com/jjy1994/BadEncoder.

Jinyuan Jia, Xiaoyu Cao, Binghui Wang et al.

It is well-known that classifiers are vulnerable to adversarial perturbations. To defend against adversarial perturbations, various certified robustness results have been derived. However, existing certified robustnesses are limited to top-1 predictions. In many real-world applications, top-$k$ predictions are more relevant. In this work, we aim to derive certified robustness for top-$k$ predictions. In particular, our certified robustness is based on randomized smoothing, which turns any classifier to a new classifier via adding noise to an input example. We adopt randomized smoothing because it is scalable to large-scale neural networks and applicable to any classifier. We derive a tight robustness in $\ell_2$ norm for top-$k$ predictions when using randomized smoothing with Gaussian noise. We find that generalizing the certified robustness from top-1 to top-$k$ predictions faces significant technical challenges. We also empirically evaluate our method on CIFAR10 and ImageNet. For example, our method can obtain an ImageNet classifier with a certified top-5 accuracy of 62.8\% when the $\ell_2$-norms of the adversarial perturbations are less than 0.5 (=127/255). Our code is publicly available at: \url{https://github.com/jjy1994/Certify_Topk}.

Lingyu Du, Yupei Liu, Jinyuan Jia et al.

Deep regression models are used in a wide variety of safety-critical applications, but are vulnerable to backdoor attacks. Although many defenses have been proposed for classification models, they are ineffective as they do not consider the uniqueness of regression models. First, the outputs of regression models are continuous values instead of discretized labels. Thus, the potential infected target of a backdoored regression model has infinite possibilities, which makes it impossible to be determined by existing defenses. Second, the backdoor behavior of backdoored deep regression models is triggered by the activation values of all the neurons in the feature space, which makes it difficult to be detected and mitigated using existing defenses. To resolve these problems, we propose DRMGuard, the first defense to identify if a deep regression model in the image domain is backdoored or not. DRMGuard formulates the optimization problem for reverse engineering based on the unique output-space and feature-space characteristics of backdoored deep regression models. We conduct extensive evaluations on two regression tasks and four datasets. The results show that DRMGuard can consistently defend against various backdoor attacks. We also generalize four state-of-the-art defenses designed for classifiers to regression models, and compare DRMGuard with them. The results show that DRMGuard significantly outperforms all those defenses.

Bochuan Cao, Jinyuan Jia, Chuxuan Hu et al.

Backdoor attacks aim to inject a backdoor into a classifier such that it predicts any input with an attacker-chosen backdoor trigger as an attacker-chosen target class. Existing backdoor attacks require either retraining the classifier with some clean data or modifying the model's architecture. As a result, they are 1) not applicable when clean data is unavailable, 2) less efficient when the model is large, and 3) less stealthy due to architecture changes. In this work, we propose DFBA, a novel retraining-free and data-free backdoor attack without changing the model architecture. Technically, our proposed method modifies a few parameters of a classifier to inject a backdoor. Through theoretical analysis, we verify that our injected backdoor is provably undetectable and unremovable by various state-of-the-art defenses under mild assumptions. Our evaluation on multiple datasets further demonstrates that our injected backdoor: 1) incurs negligible classification loss, 2) achieves 100% attack success rates, and 3) bypasses six existing state-of-the-art defenses. Moreover, our comparison with a state-of-the-art non-data-free backdoor attack shows our attack is more stealthy and effective against various defenses while achieving less classification accuracy loss.

Zhangchen Xu, Fengqing Jiang, Luyao Niu et al. · uw

Federated learning (FL) enables multiple participants to train a global machine learning model without sharing their private training data. Peer-to-peer (P2P) FL advances existing centralized FL paradigms by eliminating the server that aggregates local models from participants and then updates the global model. However, P2P FL is vulnerable to (i) honest-but-curious participants whose objective is to infer private training data of other participants, and (ii) Byzantine participants who can transmit arbitrarily manipulated local models to corrupt the learning process. P2P FL schemes that simultaneously guarantee Byzantine resilience and preserve privacy have been less studied. In this paper, we develop Brave, a protocol that ensures Byzantine Resilience And privacy-preserving property for P2P FL in the presence of both types of adversaries. We show that Brave preserves privacy by establishing that any honest-but-curious adversary cannot infer other participants' private data by observing their models. We further prove that Brave is Byzantine-resilient, which guarantees that all benign participants converge to an identical model that deviates from a global model trained without Byzantine adversaries by a bounded distance. We evaluate Brave against three state-of-the-art adversaries on a P2P FL for image classification tasks on benchmark datasets CIFAR10 and MNIST. Our results show that the global model learned with Brave in the presence of adversaries achieves comparable classification accuracy to a global model trained in the absence of any adversary.

Yupei Liu, Jinyuan Jia, Hongbin Liu et al.

Pre-trained encoders are general-purpose feature extractors that can be used for many downstream tasks. Recent progress in self-supervised learning can pre-train highly effective encoders using a large volume of unlabeled data, leading to the emerging encoder as a service (EaaS). A pre-trained encoder may be deemed confidential because its training requires lots of data and computation resources as well as its public release may facilitate misuse of AI, e.g., for deepfakes generation. In this paper, we propose the first attack called StolenEncoder to steal pre-trained image encoders. We evaluate StolenEncoder on multiple target encoders pre-trained by ourselves and three real-world target encoders including the ImageNet encoder pre-trained by Google, CLIP encoder pre-trained by OpenAI, and Clarifai's General Embedding encoder deployed as a paid EaaS. Our results show that our stolen encoders have similar functionality with the target encoders. In particular, the downstream classifiers built upon a target encoder and a stolen one have similar accuracy. Moreover, stealing a target encoder using StolenEncoder requires much less data and computation resources than pre-training it from scratch. We also explore three defenses that perturb feature vectors produced by a target encoder. Our results show these defenses are not enough to mitigate StolenEncoder.

Jinyuan Jia, Hongbin Liu, Neil Zhenqiang Gong

Foundation models--such as GPT, CLIP, and DINO--have achieved revolutionary progress in the past several years and are commonly believed to be a promising approach for general-purpose AI. In particular, self-supervised learning is adopted to pre-train a foundation model using a large amount of unlabeled data. A pre-trained foundation model is like an ``operating system'' of the AI ecosystem. Specifically, a foundation model can be used as a feature extractor for many downstream tasks with little or no labeled training data. Existing studies on foundation models mainly focused on pre-training a better foundation model to improve its performance on downstream tasks in non-adversarial settings, leaving its security and privacy in adversarial settings largely unexplored. A security or privacy issue of a pre-trained foundation model leads to a single point of failure for the AI ecosystem. In this book chapter, we discuss 10 basic security and privacy problems for the pre-trained foundation models, including six confidentiality problems, three integrity problems, and one availability problem. For each problem, we discuss potential opportunities and challenges. We hope our book chapter will inspire future research on the security and privacy of foundation models.

Hongbin Liu, Jinyuan Jia, Wenjie Qu et al.

Given a set of unlabeled images or (image, text) pairs, contrastive learning aims to pre-train an image encoder that can be used as a feature extractor for many downstream tasks. In this work, we propose EncoderMI, the first membership inference method against image encoders pre-trained by contrastive learning. In particular, given an input and a black-box access to an image encoder, EncoderMI aims to infer whether the input is in the training dataset of the image encoder. EncoderMI can be used 1) by a data owner to audit whether its (public) data was used to pre-train an image encoder without its authorization or 2) by an attacker to compromise privacy of the training data when it is private/sensitive. Our EncoderMI exploits the overfitting of the image encoder towards its training data. In particular, an overfitted image encoder is more likely to output more (or less) similar feature vectors for two augmented versions of an input in (or not in) its training dataset. We evaluate EncoderMI on image encoders pre-trained on multiple datasets by ourselves as well as the Contrastive Language-Image Pre-training (CLIP) image encoder, which is pre-trained on 400 million (image, text) pairs collected from the Internet and released by OpenAI. Our results show that EncoderMI can achieve high accuracy, precision, and recall. We also explore a countermeasure against EncoderMI via preventing overfitting through early stopping. Our results show that it achieves trade-offs between accuracy of EncoderMI and utility of the image encoder, i.e., it can reduce the accuracy of EncoderMI, but it also incurs classification accuracy loss of the downstream classifiers built based on the image encoder.

Hongbin Liu, Jinyuan Jia, Neil Zhenqiang Gong

3D point cloud classification has many safety-critical applications such as autonomous driving and robotic grasping. However, several studies showed that it is vulnerable to adversarial attacks. In particular, an attacker can make a classifier predict an incorrect label for a 3D point cloud via carefully modifying, adding, and/or deleting a small number of its points. Randomized smoothing is state-of-the-art technique to build certifiably robust 2D image classifiers. However, when applied to 3D point cloud classification, randomized smoothing can only certify robustness against adversarially modified points. In this work, we propose PointGuard, the first defense that has provable robustness guarantees against adversarially modified, added, and/or deleted points. Specifically, given a 3D point cloud and an arbitrary point cloud classifier, our PointGuard first creates multiple subsampled point clouds, each of which contains a random subset of the points in the original point cloud; then our PointGuard predicts the label of the original point cloud as the majority vote among the labels of the subsampled point clouds predicted by the point cloud classifier. Our first major theoretical contribution is that we show PointGuard provably predicts the same label for a 3D point cloud when the number of adversarially modified, added, and/or deleted points is bounded. Our second major theoretical contribution is that we prove the tightness of our derived bound when no assumptions on the point cloud classifier are made. Moreover, we design an efficient algorithm to compute our certified robustness guarantees. We also empirically evaluate PointGuard on ModelNet40 and ScanNet benchmark datasets.

Binghui Wang, Jinyuan Jia, Neil Zhenqiang Gong

Semi-supervised node classification on graph-structured data has many applications such as fraud detection, fake account and review detection, user's private attribute inference in social networks, and community detection. Various methods such as pairwise Markov Random Fields (pMRF) and graph neural networks were developed for semi-supervised node classification. pMRF is more efficient than graph neural networks. However, existing pMRF-based methods are less accurate than graph neural networks, due to a key limitation that they assume a heuristics-based constant edge potential for all edges. In this work, we aim to address the key limitation of existing pMRF-based methods. In particular, we propose to learn edge potentials for pMRF. Our evaluation results on various types of graph datasets show that our optimized pMRF-based method consistently outperforms existing graph neural networks in terms of both accuracy and efficiency. Our results highlight that previous work may have underestimated the power of pMRF for semi-supervised node classification.

Jinyuan Jia, Yupei Liu, Xiaoyu Cao et al.

Data poisoning attacks and backdoor attacks aim to corrupt a machine learning classifier via modifying, adding, and/or removing some carefully selected training examples, such that the corrupted classifier makes incorrect predictions as the attacker desires. The key idea of state-of-the-art certified defenses against data poisoning attacks and backdoor attacks is to create a majority vote mechanism to predict the label of a testing example. Moreover, each voter is a base classifier trained on a subset of the training dataset. Classical simple learning algorithms such as k nearest neighbors (kNN) and radius nearest neighbors (rNN) have intrinsic majority vote mechanisms. In this work, we show that the intrinsic majority vote mechanisms in kNN and rNN already provide certified robustness guarantees against data poisoning attacks and backdoor attacks. Moreover, our evaluation results on MNIST and CIFAR10 show that the intrinsic certified robustness guarantees of kNN and rNN outperform those provided by state-of-the-art certified defenses. Our results serve as standard baselines for future certified defenses against data poisoning attacks and backdoor attacks.

Xinlei He, Jinyuan Jia, Michael Backes et al.

Graph data, such as chemical networks and social networks, may be deemed confidential/private because the data owner often spends lots of resources collecting the data or the data contains sensitive information, e.g., social relationships. Recently, neural networks were extended to graph data, which are known as graph neural networks (GNNs). Due to their superior performance, GNNs have many applications, such as healthcare analytics, recommender systems, and fraud detection. In this work, we propose the first attacks to steal a graph from the outputs of a GNN model that is trained on the graph. Specifically, given a black-box access to a GNN model, our attacks can infer whether there exists a link between any pair of nodes in the graph used to train the model. We call our attacks link stealing attacks. We propose a threat model to systematically characterize an adversary's background knowledge along three dimensions which in total leads to a comprehensive taxonomy of 8 different link stealing attacks. We propose multiple novel methods to realize these 8 attacks. Extensive experiments on 8 real-world datasets show that our attacks are effective at stealing links, e.g., AUC (area under the ROC curve) is above 0.95 in multiple cases. Our results indicate that the outputs of a GNN model reveal rich information about the structure of the graph used to train the model.

Binghui Wang, Xiaoyu Cao, Jinyuan jia et al.

Backdoor attack is a severe security threat to deep neural networks (DNNs). We envision that, like adversarial examples, there will be a cat-and-mouse game for backdoor attacks, i.e., new empirical defenses are developed to defend against backdoor attacks but they are soon broken by strong adaptive backdoor attacks. To prevent such cat-and-mouse game, we take the first step towards certified defenses against backdoor attacks. Specifically, in this work, we study the feasibility and effectiveness of certifying robustness against backdoor attacks using a recent technique called randomized smoothing. Randomized smoothing was originally developed to certify robustness against adversarial examples. We generalize randomized smoothing to defend against backdoor attacks. Our results show the theoretical feasibility of using randomized smoothing to certify robustness against backdoor attacks. However, we also find that existing randomized smoothing methods have limited effectiveness at defending against backdoor attacks, which highlight the needs of new theory and methods to certify robustness against backdoor attacks.

Xiaoyu Cao, Jinyuan Jia, Neil Zhenqiang Gong

Local Differential Privacy (LDP) protocols enable an untrusted data collector to perform privacy-preserving data analytics. In particular, each user locally perturbs its data to preserve privacy before sending it to the data collector, who aggregates the perturbed data to obtain statistics of interest. In the past several years, researchers from multiple communities -- such as security, database, and theoretical computer science -- have proposed many LDP protocols. These studies mainly focused on improving the utility of the LDP protocols. However, the security of LDP protocols is largely unexplored. In this work, we aim to bridge this gap. We focus on LDP protocols for frequency estimation and heavy hitter identification, which are two basic data analytics tasks. Specifically, we show that an attacker can inject fake users into an LDP protocol and the fake users send carefully crafted data to the data collector such that the LDP protocol estimates high frequencies for arbitrary attacker-chosen items or identifies them as heavy hitters. We call our attacks data poisoning attacks. We theoretically and/or empirically show the effectiveness of our attacks. We also explore three countermeasures against our attacks. Our experimental results show that they can effectively defend against our attacks in some scenarios but have limited effectiveness in others, highlighting the needs for new defenses against our attacks.

Xiaoyu Cao, Jinyuan Jia, Neil Zhenqiang Gong

A deep neural network (DNN) classifier represents a model owner's intellectual property as training a DNN classifier often requires lots of resource. Watermarking was recently proposed to protect the intellectual property of DNN classifiers. However, watermarking suffers from a key limitation: it sacrifices the utility/accuracy of the model owner's classifier because it tampers the classifier's training or fine-tuning process. In this work, we propose IPGuard, the first method to protect intellectual property of DNN classifiers that provably incurs no accuracy loss for the classifiers. Our key observation is that a DNN classifier can be uniquely represented by its classification boundary. Based on this observation, IPGuard extracts some data points near the classification boundary of the model owner's classifier and uses them to fingerprint the classifier. A DNN classifier is said to be a pirated version of the model owner's classifier if they predict the same labels for most fingerprinting data points. IPGuard is qualitatively different from watermarking. Specifically, IPGuard extracts fingerprinting data points near the classification boundary of a classifier that is already trained, while watermarking embeds watermarks into a classifier during its training or fine-tuning process. We extensively evaluate IPGuard on CIFAR-10, CIFAR-100, and ImageNet datasets. Our results show that IPGuard can robustly identify post-processed versions of the model owner's classifier as pirated versions of the classifier, and IPGuard can identify classifiers, which are not the model owner's classifier nor its post-processed versions, as non-pirated versions of the classifier.

Jinyuan Jia, Ahmed Salem, Michael Backes et al.

In a membership inference attack, an attacker aims to infer whether a data sample is in a target classifier's training dataset or not. Specifically, given a black-box access to the target classifier, the attacker trains a binary classifier, which takes a data sample's confidence score vector predicted by the target classifier as an input and predicts the data sample to be a member or non-member of the target classifier's training dataset. Membership inference attacks pose severe privacy and security threats to the training dataset. Most existing defenses leverage differential privacy when training the target classifier or regularize the training process of the target classifier. These defenses suffer from two key limitations: 1) they do not have formal utility-loss guarantees of the confidence score vectors, and 2) they achieve suboptimal privacy-utility tradeoffs. In this work, we propose MemGuard, the first defense with formal utility-loss guarantees against black-box membership inference attacks. Instead of tampering the training process of the target classifier, MemGuard adds noise to each confidence score vector predicted by the target classifier. Our key observation is that attacker uses a classifier to predict member or non-member and classifier is vulnerable to adversarial examples. Based on the observation, we propose to add a carefully crafted noise vector to a confidence score vector to turn it into an adversarial example that misleads the attacker's classifier. Our experimental results on three datasets show that MemGuard can effectively defend against membership inference attacks and achieve better privacy-utility tradeoffs than existing defenses. Our work is the first one to show that adversarial examples can be used as defensive mechanisms to defend against membership inference attacks.

Binghui Wang, Jinyuan Jia, Neil Zhenqiang Gong

Many security and privacy problems can be modeled as a graph classification problem, where nodes in the graph are classified by collective classification simultaneously. State-of-the-art collective classification methods for such graph-based security and privacy analytics follow the following paradigm: assign weights to edges of the graph, iteratively propagate reputation scores of nodes among the weighted graph, and use the final reputation scores to classify nodes in the graph. The key challenge is to assign edge weights such that an edge has a large weight if the two corresponding nodes have the same label, and a small weight otherwise. Although collective classification has been studied and applied for security and privacy problems for more than a decade, how to address this challenge is still an open question. In this work, we propose a novel collective classification framework to address this long-standing challenge. We first formulate learning edge weights as an optimization problem, which quantifies the goals about the final reputation scores that we aim to achieve. However, it is computationally hard to solve the optimization problem because the final reputation scores depend on the edge weights in a very complex way. To address the computational challenge, we propose to jointly learn the edge weights and propagate the reputation scores, which is essentially an approximate solution to the optimization problem. We compare our framework with state-of-the-art methods for graph-based security and privacy analytics using four large-scale real-world datasets from various application scenarios such as Sybil detection in social networks, fake review detection in Yelp, and attribute inference attacks. Our results demonstrate that our framework achieves higher accuracies than state-of-the-art methods with an acceptable computational overhead.

Jinyuan Jia, Neil Zhenqiang Gong

Users in various web and mobile applications are vulnerable to attribute inference attacks, in which an attacker leverages a machine learning classifier to infer a target user's private attributes (e.g., location, sexual orientation, political view) from its public data (e.g., rating scores, page likes). Existing defenses leverage game theory or heuristics based on correlations between the public data and attributes. These defenses are not practical. Specifically, game-theoretic defenses require solving intractable optimization problems, while correlation-based defenses incur large utility loss of users' public data. In this paper, we present AttriGuard, a practical defense against attribute inference attacks. AttriGuard is computationally tractable and has small utility loss. Our AttriGuard works in two phases. Suppose we aim to protect a user's private attribute. In Phase I, for each value of the attribute, we find a minimum noise such that if we add the noise to the user's public data, then the attacker's classifier is very likely to infer the attribute value for the user. We find the minimum noise via adapting existing evasion attacks in adversarial machine learning. In Phase II, we sample one attribute value according to a certain probability distribution and add the corresponding noise found in Phase I to the user's public data. We formulate finding the probability distribution as solving a constrained convex optimization problem. We extensively evaluate AttriGuard and compare it with existing methods using a real-world dataset. Our results show that AttriGuard substantially outperforms existing methods. Our work is the first one that shows evasion attacks can be used as defensive techniques for privacy protection.