Evaluations of Machine Learning Privacy Defenses are Misleading

arXiv:2404.17399v1 Announce Type: new
Abstract: Empirical defenses for machine learning privacy forgo the provable guarantees of differential privacy in the hope of achieving higher utility while resisting realistic adversaries. We identify severe pitfalls in existing empirical privacy evaluations (based on membership inference attacks) that result in misleading conclusions. In particular, we show that prior evaluations fail to characterize the privacy leakage of the most vulnerable samples, use weak attacks, and avoid comparisons with practical differential privacy baselines. In 5 case …

adversaries arxiv attacks cs.cr cs.lg defenses differential privacy higher hope identify machine machine learning privacy result utility