PANORAMIA: Privacy Auditing of Machine Learning Models without Retraining

Mishaal Kazmi, Hadrien Lautraite, Alireza Akbari, Mauricio Soroco, Qiaoyue Tang, Tao Wang, Sébastien Gambs, Mathias Lécuyer

Abstract: We introduce a privacy auditing scheme for ML models that relies on membership inference attacks using generated data as "non-members". This scheme, which we call PANORAMIA, quantifies the privacy leakage for large-scale ML models without control of the training process or model re-training and only requires access to a subset of the training data. To demonstrate its applicability, we evaluate our auditing scheme across multiple ML domains, ranging from image and tabular data classification to large-scale language models.