Membership inference attacks (MIAs) are used to test practical privacy of machine learning models. MIAs complement formal guarantees from differential privacy (DP) under a more realistic adversary model. We analyze MIA vulnerability of fine-tuned neural networks both empirically and theoretically, the latter using a simplified model of fine-tuning. We show that the vulnerability of non-DP models when measured as the attacker advantage at a fixed false positive rate reduces according to a simple power law as the number of examples per class increases. A similar power-law applies even for the most vulnerable points, but the dataset size needed for adequate protection of the most vulnerable points is very large.

Federated Learning (FL) has emerged as a promising paradigm for collaborative model training without the need to share clients' personal data, thereby preserving privacy. However, the non-IID nature of the clients' data introduces major challenges for FL, highlighting the importance of personalized federated learning (PFL) methods. In PFL, models are trained to cater to specific feature distributions present in the population data. A notable method for PFL is the Iterative Federated Clustering Algorithm (IFCA), which mitigates the concerns associated with the non-IID-ness by grouping clients with similar data distributions. While it has been shown that IFCA enhances both accuracy and fairness, its strategy of dividing the population into smaller clusters increases vulnerability to Membership Inference Attacks (MIA), particularly among minorities with limited training samples. In this paper, we introduce IFCA-MIR, an improved version of IFCA that integrates MIA risk assessment into the clustering process. Allowing clients to select clusters based on both model performance and MIA vulnerability, IFCA-MIR achieves an improved performance with respect to accuracy, fairness, and privacy. We demonstrate that IFCA-MIR significantly reduces MIA risk while maintaining comparable model accuracy and fairness as the original IFCA.

As machine learning models increasingly process sensitive data, understanding their vulnerability to privacy attacks is vital. Membership inference attacks (MIAs) exploit model responses to infer whether specific data points were used during training, posing a significant privacy risk. Prior research suggests that spiking neural networks (SNNs), which rely on event-driven computation and discrete spike-based encoding, exhibit greater resilience to MIAs than artificial neural networks (ANNs). This resilience stems from their non-differentiable activations and inherent stochasticity, which obscure the correlation between model responses and individual training samples. To enhance privacy in SNNs, we explore two techniques: quantization and surrogate gradients. Quantization, which reduces precision to limit information leakage, has improved privacy in ANNs. Given SNNs' sparse and irregular activations, quantization may further disrupt the activation patterns exploited by MIAs. We assess the vulnerability of SNNs and ANNs under weight and activation quantization across multiple datasets, using the attack model's receiver operating characteristic (ROC) curve area under the curve (AUC) metric, where lower values indicate stronger privacy, and evaluate the privacy-accuracy trade-off. Our findings show that quantization enhances privacy in both architectures with minimal performance loss, though full-precision SNNs remain more resilient than quantized ANNs. Additionally, we examine the impact of surrogate gradients on privacy in SNNs. Among five evaluated gradients, spike rate escape provides the best privacy-accuracy trade-off, while arctangent increases vulnerability to MIAs. These results reinforce SNNs' inherent privacy advantages and demonstrate that quantization and surrogate gradient selection significantly influence privacy-accuracy trade-offs in SNNs.

Membership Inference Attacks (MIAs) have emerged as a valuable framework for evaluating privacy leakage by machine learning models. Score-based MIAs are distinguished, in particular, by their ability to exploit the confidence scores that the model generates for particular inputs. Existing score-based MIAs implicitly assume that the adversary has access to the target model's hyperparameters, which can be used to train the shadow models for the attack. In this work, we demonstrate that the knowledge of target hyperparameters is not a prerequisite for MIA in the transfer learning setting. Based on this, we propose a novel approach to select the hyperparameters for training the shadow models for MIA when the attacker has no prior knowledge about them by matching the output distributions of target and shadow models. We demonstrate that using the new approach yields hyperparameters that lead to an attack near indistinguishable in performance from an attack that uses target hyperparameters to train the shadow models. Furthermore, we study the empirical privacy risk of unaccounted use of training data for hyperparameter optimization (HPO) in differentially private (DP) transfer learning. We find no statistically significant evidence that performing HPO using training data would increase vulnerability to MIA.

The primary promise of decentralized learning is to allow users to engage in the training of machine learning models in a collaborative manner while keeping their data on their premises and without relying on any central entity. However, this paradigm necessitates the exchange of model parameters or gradients between peers. Such exchanges can be exploited to infer sensitive information about training data, which is achieved through privacy attacks (e.g Membership Inference Attacks -- MIA). In order to devise effective defense mechanisms, it is important to understand the factors that increase/reduce the vulnerability of a given decentralized learning architecture to MIA. In this study, we extensively explore the vulnerability to MIA of various decentralized learning architectures by varying the graph structure (e.g number of neighbors), the graph dynamics, and the aggregation strategy, across diverse datasets and data distributions. Our key finding, which to the best of our knowledge we are the first to report, is that the vulnerability to MIA is heavily correlated to (i) the local model mixing strategy performed by each node upon reception of models from neighboring nodes and (ii) the global mixing properties of the communication graph. We illustrate these results experimentally using four datasets and by theoretically analyzing the mixing properties of various decentralized architectures. Our paper draws a set of lessons learned for devising decentralized learning systems that reduce by design the vulnerability to MIA.

We apply a state-of-the-art membership inference attack (MIA) to systematically test the practical privacy vulnerability of fine-tuning large image classification models.We focus on understanding the properties of data sets and samples that make them vulnerable to membership inference. In terms of data set properties, we find a strong power law dependence between the number of examples per class in the data and the MIA vulnerability, as measured by true positive rate of the attack at a low false positive rate. For an individual sample, large gradients at the end of training are strongly correlated with MIA vulnerability.