Bounding Reconstruction Attack Success of Adversaries Without Data Priors
Authors:
Alexander Ziller,
Anneliese Riess,
Kristian Schwethelm,
Tamara T. Mueller,
Daniel Rueckert,
Georgios Kaissis
Abstract:
Reconstruction attacks on machine learning (ML) models pose a strong risk of leakage of sensitive data. In specific contexts, an adversary can (almost) perfectly reconstruct training data samples from a trained model using the model's gradients. When training ML models with differential privacy (DP), formal upper bounds on the success of such reconstruction attacks can be provided. So far, these b…
▽ More
Reconstruction attacks on machine learning (ML) models pose a strong risk of leakage of sensitive data. In specific contexts, an adversary can (almost) perfectly reconstruct training data samples from a trained model using the model's gradients. When training ML models with differential privacy (DP), formal upper bounds on the success of such reconstruction attacks can be provided. So far, these bounds have been formulated under worst-case assumptions that might not hold high realistic practicality. In this work, we provide formal upper bounds on reconstruction success under realistic adversarial settings against ML models trained with DP and support these bounds with empirical results. With this, we show that in realistic scenarios, (a) the expected reconstruction success can be bounded appropriately in different contexts and by different metrics, which (b) allows for a more educated choice of a privacy parameter.
△ Less
Submitted 20 February, 2024;
originally announced February 2024.
