View PDF HTML (experimental)

Abstract:While Federated Learning (FL) mitigates direct data exposure, the resulting trained models remain susceptible to membership inference attacks (MIAs). This paper presents an empirical evaluation of Differential Privacy (DP) as a defense mechanism against MIAs in FL, leveraging the environment of the 2025 NIST Genomics Privacy-Preserving Federated Learning (PPFL) Red Teaming Event. To improve inference accuracy, we propose a stacking attack strategy that ensembles seven black-box estimators to train a meta-classifier on prediction probabilities and cross-entropy losses. We evaluate this methodology against target models under three privacy configurations: an unprotected convolutional neural network (CNN, $\epsilon=\infty$), a low-privacy DP model ($\epsilon=200$), and a high-privacy DP model ($\epsilon=10$). The attack outperforms all baselines in the No DP and Low Privacy settings and, critically, maintains measurable membership leakage at $\epsilon=200$ where a single-signal LiRA baseline collapses. Evaluated on an independent third-party benchmark, these results provide an empirical characterisation of how stacking-based inference degrades across calibrated DP tiers in FL.

Submission history

From: Gustavo De Carvalho Bertoli [view email]
[v1] Tue, 14 Apr 2026 13:51:02 UTC (237 KB)
[v2] Wed, 15 Apr 2026 10:00:01 UTC (237 KB)