Organizations: Department of Computer and Data Sciences, Case School of Engineering, Case Western Reserve University, Cleveland, OH, USA · Department of Bioethics, School of Medicine, Case Western Reserve University, Cleveland, OH, USA
Genomic data sharing is essential for biomedical research but introduces significant privacy risks, including membership inference attacks, in which an adversary determines whether a specific individual's data was used in a study. Existing privacy protections are typically applied after a release format has already been chosen, leaving researchers without a principled method to assess dataset-specific risk beforehand. We propose a server-assisted, principal component analysis-based pre-release auditing framework that maps genomic data into a shared low-dimensional space, applies epsilon-local differential privacy via the Laplace mechanism with a formally derived sensitivity for principal component analysis scores, and estimates membership inference attack power at a controlled 5% false-positive rate using a distance-threshold attack. A membership-neutrality diagnostic based on Unbiased Reconstruction Robustness is reported alongside attack power, precision, and recall. Experiments using the openSNP and 1000 Genomes Project datasets across multiple dimensionalities and privacy budgets show that separation between the researcher's dataset and public auxiliary data in principal component analysis space is a dominant determinant of attack success, exceeding the effect of perturbation level alone. Comparison against a genome-wide association study-style likelihood ratio test baseline further contextualizes the measured risk. This framework enables researchers, data repositories, and institutional review bodies to quantify dataset-specific membership inference risk before genomic data publication, supporting more informed and privacy-aware biomedical data-sharing decisions.
Figures & tables
Symbol
Description
DR
Researcher’s Dataset
NDR
Researcher’s noisy data produced as a result of Laplace noise
Dp
Publicly available dataset
MIA
Membership Inference Attack
ϵ , eps
Privacy Parameter
k
Number of clusters
TABLE I : Table of symbols and notations.
Fig. 1 : An overview of the proposed framework. (1) The researcher collects genomic data, performs local PCA, shares the PCA model and DP-noised scores with the server. (2) The server integrates public datasets and, using the shared PCA model, maps auxiliaries into the same coordinate system, and (3) The server estimates membership inference risk by analyzing the spatial relationships among the datasets. (4) The resulting risk estimation is reported back to the researcher, guiding them in fine-tuning privacy parameters prior to publishing research outcomes.
# released SNPs L
Power (Bernoulli LRT)
10
0.15
100
0.13
1,000
0.39
5,000
0.59
10,000
0.60
28,000
0.59
TABLE II : LRT baseline on openSNP (study size n=100 , 50 cases / 50 controls using the real data; reference = remaining individuals), power at FPR=5% . SNPs are selected by GWAS p -value computed on the study and released as MAFs.
Fig. 2 : Impact of centroid distances on membership inference attack metrics under varying privacy budgets ( ϵ ) for the openSNP and 1000 Genomes datasets. Panels (a–c) show power, recall, and precision for openSNP, respectively; panels (d–f) show the corresponding metrics for 1000 Genomes.
Fig. 3 : Membership inference attack (MIA) power under different distance thresholds and privacy budgets for the openSNP and 1000 Genomes datasets.
The use of synthetic data has become increasingly popular as a privacy-preserving alternative to sharing real datasets, especially in sensitive domains such as healthcare, finance, and demography. However, the privacy assurances of synthetic data are not absolute, and remain susceptible to membership inference attacks (MIAs), where adversaries aim to determine whether a specific individual was present in the dataset used to train the generator. In this work, we propose a practical and effective method to quantify membership disclosure risk in tabular synthetic datasets using kernel density estimators (KDEs). Our KDE-based approach models the distribution of nearest-neighbour distances between synthetic data and the training records, allowing probabilistic inference of membership and enabling robust evaluation via ROC curves. We propose two attack models: a 'True Distribution Attack', which assumes privileged access to training data, and a more realistic, implementable 'Realistic Attack' that uses auxiliary data without true membership labels. Empirical evaluations across four real-world datasets and six synthetic data generators demonstrate that our method consistently achieves higher F1 scores and sharper risk characterization than a prior baseline approach, without requiring computationally expensive shadow models. The proposed method provides a practical framework and metric for quantifying membership disclosure risk in synthetic data, which enables data custodians to conduct a post-generation risk assessment prior to releasing their synthetic datasets for downstream use. The datasets and codes for this study are available at https://github.com/PyCoder913/MIA-KDE.
Rajdeep Pathak, Amit Basak, Sayantee Jana
SCAI, Sorbonne Université, Paris, France · SAFIR, Sorbonne University Abu Dhabi, UAE · Indian Institute of Technology Hyderabad, India
Membership inference attacks (MIAs) aim to determine whether a data sample was included in a machine learning (ML) model's training set and have become the de facto standard for measuring privacy leakages in ML. We propose an evaluation framework that defines the conditions under which MIAs constitute a genuine privacy threat, and review representative MIAs against it. We find that, under the realistic conditions defined in our framework, MIAs represent weak privacy threats. Thus, relying on them as a privacy metric in ML can lead to an overestimation of risk and to unnecessary sacrifices in model utility as a consequence of employing too strong defenses.
Najeeb Jebreel, David Sánchez, Josep Domingo-Ferrer
Universitat Rovira i Virgili, Department of Computer Engineering and Mathematics, CYBERCAT-Center for Cybersecurity Research of Catalonia, ComSCIAM-Center for Computational Science and Applied Mathematics Av. Països Catalans 26, 43007 Tarragona, Catalonia
Tabular data sharing under privacy constraints is increasingly important for research and collaboration. Synthetic data generators (SDGs) are a promising solution, but synthetic data remains vulnerable to attacks, such as membership inference attacks (MIAs), which aim to determine whether a specific record was part of the training data. State-of-the-art MIAs are powerful but impractical: they rely on shadow modeling, requiring hundreds of SDG training runs, and need auxiliary data several times larger than the original training set. Fast proxy metrics like distance to closest record (DCR) are efficient but have limited sensitivity to MIA risk. We introduce ReMIA (Relative Membership Inference Attack), a practical privacy metric that requires only two SDG training runs and additional data no larger than the original training set. Rather than predicting whether a record was in the training set, ReMIA generates two synthetic datasets from two source datasets and measures whether a classifier can identify which source a record came from. Experiments across multiple tabular datasets and SDGs show that ReMIA has a sensitivity comparable to state-of-the-art MIAs while being substantially more practical. We further observe that SDGs can achieve privacy-utility trade-offs that traditional noise-based anonymization methods do not match. Code is available at https://github.com/aindo-com/remia.