The marriage of federated learning and recommender system (FedRec) has been widely used to address the growing data privacy concerns in personalized recommendation services. In FedRecs, users' attribute information and behavior data (i.e., user-item interaction data) are kept locally on their personal devices, therefore, it is considered a fairly secure approach to protect user privacy. As a result, the privacy issue of FedRecs is rarely explored. Unfortunately, several recent studies reveal that FedRecs are vulnerable to user attribute inference attacks, highlighting the privacy concerns of FedRecs. In this paper, we further investigate the privacy problem of user behavior data (i.e., user-item interactions) in FedRecs. Specifically, we perform the first systematic study on interaction-level membership inference attacks on FedRecs. An interaction-level membership inference attacker is first designed, and then the classical privacy protection mechanism, Local Differential Privacy (LDP), is adopted to defend against the membership inference attack. Unfortunately, the empirical analysis shows that LDP is not effective against such new attacks unless the recommendation performance is largely compromised. To mitigate the interaction-level membership attack threats, we design a simple yet effective defense method to significantly reduce the attacker's inference accuracy without losing recommendation performance. Extensive experiments are conducted with two widely used FedRecs (Fed-NCF and Fed-LightGCN) on three real-world recommendation datasets (MovieLens-100K, Steam-200K, and Amazon Cell Phone), and the experimental results show the effectiveness of our solutions.
Recent advances in synthetic data generation (SDG) have been hailed as a solution to the difficult problem of sharing sensitive data while protecting privacy. SDG aims to learn statistical properties of real data in order to generate "artificial" data that are structurally and statistically similar to sensitive data. However, prior research suggests that inference attacks on synthetic data can undermine privacy, but only for specific outlier records. In this work, we introduce a new attribute inference attack against synthetic data. The attack is based on linear reconstruction methods for aggregate statistics, which target all records in the dataset, not only outliers. We evaluate our attack on state-of-the-art SDG algorithms, including Probabilistic Graphical Models, Generative Adversarial Networks, and recent differentially private SDG mechanisms. By defining a formal privacy game, we show that our attack can be highly accurate even on arbitrary records, and that this is the result of individual information leakage (as opposed to population-level inference). We then systematically evaluate the tradeoff between protecting privacy and preserving statistical utility. Our findings suggest that current SDG methods cannot consistently provide sufficient privacy protection against inference attacks while retaining reasonable utility. The best method evaluated, a differentially private SDG mechanism, can provide both protection against inference attacks and reasonable utility, but only in very specific settings. Lastly, we show that releasing a larger number of synthetic records can improve utility but at the cost of making attacks far more effective.
As a crucial building block in vertical Federated Learning (vFL), Split Learning (SL) has demonstrated its practice in the two-party model training collaboration, where one party holds the features of data samples and another party holds the corresponding labels. Such method is claimed to be private considering the shared information is only the embedding vectors and gradients instead of private raw data and labels. However, some recent works have shown that the private labels could be leaked by the gradients. These existing attack only works under the classification setting where the private labels are discrete. In this work, we step further to study the leakage in the scenario of the regression model, where the private labels are continuous numbers (instead of discrete labels in classification). This makes previous attacks harder to infer the continuous labels due to the unbounded output range. To address the limitation, we propose a novel learning-based attack that integrates gradient information and extra learning regularization objectives in aspects of model training properties, which can infer the labels under regression settings effectively. The comprehensive experiments on various datasets and models have demonstrated the effectiveness of our proposed attack. We hope our work can pave the way for future analyses that make the vFL framework more secure.
With the development of information science and technology, various industries have generated massive amounts of data, and machine learning is widely used in the analysis of big data. However, if the privacy of machine learning applications' customers cannot be guaranteed, it will cause security threats and losses to users' personal privacy information and service providers. Therefore, the issue of privacy protection of machine learning has received wide attention. For centralized machine learning models, we evaluate the impact of different neural network layers, gradient, gradient norm, and fine-tuned models on member inference attack performance with prior knowledge; For the federated learning model, we discuss the location of the attacker in the target model and its attack mode. The results show that the centralized machine learning model shows more serious member information leakage in all aspects, and the accuracy of the attacker in the central parameter server is significantly higher than the local Inference attacks as participants.
Rankings are widely collected in various real-life scenarios, leading to the leakage of personal information such as users' preferences on videos or news. To protect rankings, existing works mainly develop privacy protection on a single ranking within a set of ranking or pairwise comparisons of a ranking under the $\epsilon$-differential privacy. This paper proposes a novel notion called $\epsilon$-ranking differential privacy for protecting ranks. We establish the connection between the Mallows model (Mallows, 1957) and the proposed $\epsilon$-ranking differential privacy. This allows us to develop a multistage ranking algorithm to generate synthetic rankings while satisfying the developed $\epsilon$-ranking differential privacy. Theoretical results regarding the utility of synthetic rankings in the downstream tasks, including the inference attack and the personalized ranking tasks, are established. For the inference attack, we quantify how $\epsilon$ affects the estimation of the true ranking based on synthetic rankings. For the personalized ranking task, we consider varying privacy preferences among users and quantify how their privacy preferences affect the consistency in estimating the optimal ranking function. Extensive numerical experiments are carried out to verify the theoretical results and demonstrate the effectiveness of the proposed synthetic ranking algorithm.
As machine learning expanding application, there are more and more unignorable privacy and safety issues. Especially inference attacks against Machine Learning models allow adversaries to infer sensitive information about the target model, such as training data, model parameters, etc. Inference attacks can lead to serious consequences, including violating individuals privacy, compromising the intellectual property of the owner of the machine learning model. As far as concerned, researchers have studied and analyzed in depth several types of inference attacks, albeit in isolation, but there is still a lack of a holistic rick assessment of inference attacks against machine learning models, such as their application in different scenarios, the common factors affecting the performance of these attacks and the relationship among the attacks. As a result, this paper performs a holistic risk assessment of different inference attacks against Machine Learning models. This paper focuses on three kinds of representative attacks: membership inference attack, attribute inference attack and model stealing attack. And a threat model taxonomy is established. A total of 12 target models using three model architectures, including AlexNet, ResNet18 and Simple CNN, are trained on four datasets, namely CelebA, UTKFace, STL10 and FMNIST.
Model-based attacks can infer training data information from deep neural network models. These attacks heavily depend on the attacker's knowledge of the application domain, e.g., using it to determine the auxiliary data for model-inversion attacks. However, attackers may not know what the model is used for in practice. We propose a generative adversarial network (GAN) based method to explore likely or similar domains of a target model -- the model domain inference (MDI) attack. For a given target (classification) model, we assume that the attacker knows nothing but the input and output formats and can use the model to derive the prediction for any input in the desired form. Our basic idea is to use the target model to affect a GAN training process for a candidate domain's dataset that is easy to obtain. We find that the target model may distract the training procedure less if the domain is more similar to the target domain. We then measure the distraction level with the distance between GAN-generated datasets, which can be used to rank candidate domains for the target model. Our experiments show that the auxiliary dataset from an MDI top-ranked domain can effectively boost the result of model-inversion attacks.
A distribution inference attack aims to infer statistical properties of data used to train machine learning models. These attacks are sometimes surprisingly potent, but the factors that impact distribution inference risk are not well understood and demonstrated attacks often rely on strong and unrealistic assumptions such as full knowledge of training environments even in supposedly black-box threat scenarios. To improve understanding of distribution inference risks, we develop a new black-box attack that even outperforms the best known white-box attack in most settings. Using this new attack, we evaluate distribution inference risk while relaxing a variety of assumptions about the adversary's knowledge under black-box access, like known model architectures and label-only access. Finally, we evaluate the effectiveness of previously proposed defenses and introduce new defenses. We find that although noise-based defenses appear to be ineffective, a simple re-sampling defense can be highly effective. Code is available at https://github.com/iamgroot42/dissecting_distribution_inference
We propose a framework in which multiple entities collaborate to build a machine learning model while preserving privacy of their data. The approach utilizes feature embeddings from shared/per-entity feature extractors transforming data into a feature space for cooperation between entities. We propose two specific methods and compare them with a baseline method. In Shared Feature Extractor (SFE) Learning, the entities use a shared feature extractor to compute feature embeddings of samples. In Locally Trained Feature Extractor (LTFE) Learning, each entity uses a separate feature extractor and models are trained using concatenated features from all entities. As a baseline, in Cooperatively Trained Feature Extractor (CTFE) Learning, the entities train models by sharing raw data. Secure multi-party algorithms are utilized to train models without revealing data or features in plain text. We investigate the trade-offs among SFE, LTFE, and CTFE in regard to performance, privacy leakage (using an off-the-shelf membership inference attack), and computational cost. LTFE provides the most privacy, followed by SFE, and then CTFE. Computational cost is lowest for SFE and the relative speed of CTFE and LTFE depends on network architecture. CTFE and LTFE provide the best accuracy. We use MNIST, a synthetic dataset, and a credit card fraud detection dataset for evaluations.
Neural networks are susceptible to data inference attacks such as the membership inference attack, the adversarial model inversion attack and the attribute inference attack, where the attacker could infer useful information such as the membership, the reconstruction or the sensitive attributes of a data sample from the confidence scores predicted by the target classifier. In this paper, we propose a method, namely PURIFIER, to defend against membership inference attacks. It transforms the confidence score vectors predicted by the target classifier and makes purified confidence scores indistinguishable in individual shape, statistical distribution and prediction label between members and non-members. The experimental results show that PURIFIER helps defend membership inference attacks with high effectiveness and efficiency, outperforming previous defense methods, and also incurs negligible utility loss. Besides, our further experiments show that PURIFIER is also effective in defending adversarial model inversion attacks and attribute inference attacks. For example, the inversion error is raised about 4+ times on the Facescrub530 classifier, and the attribute inference accuracy drops significantly when PURIFIER is deployed in our experiment.
It is a growing direction to utilize unintended memorization in ML models to benefit real-world applications, with recent efforts like user auditing, dataset ownership inference and forgotten data measurement. Standing on the point of ML model development, we introduce a process named data origin inference, to assist ML developers in locating missed or faulty data origin in training set without maintaining strenuous metadata. We formally define the data origin and the data origin inference task in the development of the ML model (mainly neural networks). Then we propose a novel inference strategy combining embedded-space multiple instance classification and shadow training. Diverse use cases cover language, visual and structured data, with various kinds of data origin (e.g. business, county, movie, mobile user, text author). A comprehensive performance analysis of our proposed strategy contains referenced target model layers, available testing data for each origin, and in shadow training, the implementations of feature extraction as well as shadow models. Our best inference accuracy achieves 98.96% in the language use case when the target model is a transformer-based deep neural network. Furthermore, we give a statistical analysis of different kinds of data origin to investigate what kind of origin is probably to be inferred correctly.
Group fairness and privacy are fundamental aspects in designing trustworthy machine learning models. Previous research has highlighted conflicts between group fairness and different privacy notions. We are the first to demonstrate the alignment of group fairness with the specific privacy notion of attribute privacy in a blackbox setting. Attribute privacy, quantified by the resistance to attribute inference attacks (AIAs), requires indistinguishability in the target model's output predictions. Group fairness guarantees this thereby mitigating AIAs and achieving attribute privacy. To demonstrate this, we first introduce AdaptAIA, an enhancement of existing AIAs, tailored for real-world datasets with class imbalances in sensitive attributes. Through theoretical and extensive empirical analyses, we demonstrate the efficacy of two standard group fairness algorithms (i.e., adversarial debiasing and exponentiated gradient descent) against AdaptAIA. Additionally, since using group fairness results in attribute privacy, it acts as a defense against AIAs, which is currently lacking. Overall, we show that group fairness aligns with attribute privacy at no additional cost other than the already existing trade-off with model utility.
Property inference attacks against machine learning (ML) models aim to infer properties of the training data that are unrelated to the primary task of the model, and have so far been formulated as binary decision problems, i.e., whether or not the training data have a certain property. However, in industrial and healthcare applications, the proportion of labels in the training data is quite often also considered sensitive information. In this paper we introduce a new type of property inference attack that unlike binary decision problems in literature, aim at inferring the class label distribution of the training data from parameters of ML classifier models. We propose a method based on \emph{shadow training} and a \emph{meta-classifier} trained on the parameters of the shadow classifiers augmented with the accuracy of the classifiers on auxiliary data. We evaluate the proposed approach for ML classifiers with fully connected neural network architectures. We find that the proposed \emph{meta-classifier} attack provides a maximum relative improvement of $52\%$ over state of the art.
In this paper, we initiate the study of local model reconstruction attacks for federated learning, where a honest-but-curious adversary eavesdrops the messages exchanged between a targeted client and the server, and then reconstructs the local/personalized model of the victim. The local model reconstruction attack allows the adversary to trigger other classical attacks in a more effective way, since the local model only depends on the client's data and can leak more private information than the global model learned by the server. Additionally, we propose a novel model-based attribute inference attack in federated learning leveraging the local model reconstruction attack. We provide an analytical lower-bound for this attribute inference attack. Empirical results using real world datasets confirm that our local reconstruction attack works well for both regression and classification tasks. Moreover, we benchmark our novel attribute inference attack against the state-of-the-art attacks in federated learning. Our attack results in higher reconstruction accuracy especially when the clients' datasets are heterogeneous. Our work provides a new angle for designing powerful and explainable attacks to effectively quantify the privacy risk in FL.
With the introduction of data protection and privacy regulations, it has become crucial to remove the lineage of data on demand from a machine learning (ML) model. In the last few years, there have been notable developments in machine unlearning to remove the information of certain training data efficiently and effectively from ML models. In this work, we explore unlearning for the regression problem, particularly in deep learning models. Unlearning in classification and simple linear regression has been considerably investigated. However, unlearning in deep regression models largely remains an untouched problem till now. In this work, we introduce deep regression unlearning methods that generalize well and are robust to privacy attacks. We propose the Blindspot unlearning method which uses a novel weight optimization process. A randomly initialized model, partially exposed to the retain samples and a copy of the original model are used together to selectively imprint knowledge about the data that we wish to keep and scrub off the information of the data we wish to forget. We also propose a Gaussian fine tuning method for regression unlearning. The existing unlearning metrics for classification are not directly applicable to regression unlearning. Therefore, we adapt these metrics for the regression setting. We conduct regression unlearning experiments for computer vision, natural language processing and forecasting applications. Our methods show excellent performance for all these datasets across all the metrics. Source code: https://github.com/ayu987/deep-regression-unlearning
Unlike traditional static deep neural networks (DNNs), dynamic neural networks (NNs) adjust their structures or parameters to different inputs to guarantee accuracy and computational efficiency. Meanwhile, it has been an emerging research area in deep learning recently. Although traditional static DNNs are vulnerable to the membership inference attack (MIA) , which aims to infer whether a particular point was used to train the model, little is known about how such an attack performs on the dynamic NNs. In this paper, we propose a novel MI attack against dynamic NNs, leveraging the unique policy networks mechanism of dynamic NNs to increase the effectiveness of membership inference. We conducted extensive experiments using two dynamic NNs, i.e., GaterNet, BlockDrop, on four mainstream image classification tasks, i.e., CIFAR-10, CIFAR-100, STL-10, and GTSRB. The evaluation results demonstrate that the control-flow information can significantly promote the MIA. Based on backbone-finetuning and information-fusion, our method achieves better results than baseline attack and traditional attack using intermediate information.
Authentication systems are vulnerable to model inversion attacks where an adversary is able to approximate the inverse of a target machine learning model. Biometric models are a prime candidate for this type of attack. This is because inverting a biometric model allows the attacker to produce a realistic biometric input to spoof biometric authentication systems. One of the main constraints in conducting a successful model inversion attack is the amount of training data required. In this work, we focus on iris and facial biometric systems and propose a new technique that drastically reduces the amount of training data necessary. By leveraging the output of multiple models, we are able to conduct model inversion attacks with 1/10th the training set size of Ahmad and Fuller (IJCB 2020) for iris data and 1/1000th the training set size of Mai et al. (Pattern Analysis and Machine Intelligence 2019) for facial data. We denote our new attack technique as structured random with alignment loss. Our attacks are black-box, requiring no knowledge of the weights of the target neural network, only the dimension, and values of the output vector. To show the versatility of the alignment loss, we apply our attack framework to the task of membership inference (Shokri et al., IEEE S&P 2017) on biometric data. For the iris, membership inference attack against classification networks improves from 52% to 62% accuracy.
With the development of machine learning techniques, the attention of research has been moved from single-modal learning to multi-modal learning, as real-world data exist in the form of different modalities. However, multi-modal models often carry more information than single-modal models and they are usually applied in sensitive scenarios, such as medical report generation or disease identification. Compared with the existing membership inference against machine learning classifiers, we focus on the problem that the input and output of the multi-modal models are in different modalities, such as image captioning. This work studies the privacy leakage of multi-modal models through the lens of membership inference attack, a process of determining whether a data record involves in the model training process or not. To achieve this, we propose Multi-modal Models Membership Inference (M^4I) with two attack methods to infer the membership status, named metric-based (MB) M^4I and feature-based (FB) M^4I, respectively. More specifically, MB M^4I adopts similarity metrics while attacking to infer target data membership. FB M^4I uses a pre-trained shadow multi-modal feature extractor to achieve the purpose of data inference attack by comparing the similarities from extracted input and output features. Extensive experimental results show that both attack methods can achieve strong performances. Respectively, 72.5% and 94.83% of attack success rates on average can be obtained under unrestricted scenarios. Moreover, we evaluate multiple defense mechanisms against our attacks. The source code of M^4I attacks is publicly available at https://github.com/MultimodalMI/Multimodal-membership-inference.git.
With the rise of deep learning in various applications, privacy concerns around the protection of training data have become a critical area of research. Whereas prior studies have focused on privacy risks in single-modal models, we introduce a novel method to assess privacy for multi-modal models, specifically vision-language models like CLIP. The proposed Identity Inference Attack (IDIA) reveals whether an individual was included in the training data by querying the model with images of the same person. Letting the model choose from a wide variety of possible text labels, the model reveals whether it recognizes the person and, therefore, was used for training. Our large-scale experiments on CLIP demonstrate that individuals used for training can be identified with very high accuracy. We confirm that the model has learned to associate names with depicted individuals, implying the existence of sensitive information that can be extracted by adversaries. Our results highlight the need for stronger privacy protection in large-scale models and suggest that IDIAs can be used to prove the unauthorized use of data for training and to enforce privacy laws.
Models can expose sensitive information about their training data. In an attribute inference attack, an adversary has partial knowledge of some training records and access to a model trained on those records, and infers the unknown values of a sensitive feature of those records. We study a fine-grained variant of attribute inference we call \emph{sensitive value inference}, where the adversary's goal is to identify with high confidence some records from a candidate set where the unknown attribute has a particular sensitive value. We explicitly compare attribute inference with data imputation that captures the training distribution statistics, under various assumptions about the training data available to the adversary. Our main conclusions are: (1) previous attribute inference methods do not reveal more about the training data from the model than can be inferred by an adversary without access to the trained model, but with the same knowledge of the underlying distribution as needed to train the attribute inference attack; (2) black-box attribute inference attacks rarely learn anything that cannot be learned without the model; but (3) white-box attacks, which we introduce and evaluate in the paper, can reliably identify some records with the sensitive value attribute that would not be predicted without having access to the model. Furthermore, we show that proposed defenses such as differentially private training and removing vulnerable records from training do not mitigate this privacy risk. The code for our experiments is available at \url{https://github.com/bargavj/EvaluatingDPML}.