Quantifying the Privacy of Counterfactuals by Leveraging Membership Inference Attacks Against Synthetic Data
On the map: LLM safety and security → Membership inference attacks
Abstract (arXiv)
Counterfactuals are typically used in high-stakes decision areas to explain a machine learning model by showing how changes to the user profiles result in the desired outcome. However, explaining the model's decisions through counterfactuals can also be exploited by an adversary to conduct privacy attacks against the model or its training data. Drawing on the analogy that counterfactuals provide realistic substitutes for real training data, similar to synthetic data, we demonstrate in this paper how it is possible to successfully perform privacy attacks on counterfactuals by drawing on the attacks developed against synthetic data. More precisely, we investigate the effectiveness of the membership inference attacks designed for synthetic data on various types of counterfactuals. Additionally, while existing membership inference attacks against counterfactuals usually require to be able to query the model, we show how it is possible to perform successful membership inference attacks using only a set of counterfactuals, with no access to the model from which they are generated. Our results demonstrate that model developers should be more cautious when releasing counterfactuals to various users, as it can lead to a privacy breach.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Membership inference attacks, one of 12 directions in LLM safety and security. Builds on 24 of its 46 references that are in the map. First posted on arXiv 4 Jun 2026.
Builds on
24 of 46 references.
Closest ancestors · 4
Same direction, newest first
- 2025Ensembling Membership Inference Attacks Against Tabular Generative Models2 Sep
- 2024Impact of Dataset Properties on Membership Inference Vulnerability of Deep Transfer Learning7 Feb
- 2023Membership Inference Attacks against Synthetic Data through Overfitting Detection24 Feb
- 2022On the Privacy Risks of Algorithmic Recourse10 Nov
Other references · 20
- 2016Membership Inference Attacks against Machine Learning Modelscited by 2,221
- 2017Counterfactual Explanations without Opening the Black Box: Automated Decisions and the GDPRcited by 1,086
- 2017ZOO: Zeroth Order Optimization based Black-box Attacks to Deep Neural Networks without Training Substitute Modelscited by 656
- 2019Explaining Machine Learning Classifiers through Diverse Counterfactual Explanationscited by 612
- 2019Reconstruction and Membership Inference Attacks against Generative Modelscited by 122
- 2019Synthetic Data for Deep Learningcited by 113
- 2021Winning the NIST Contest: A scalable and general approach to differentially private synthetic datacited by 101
- 2019The Dangers of Post-hoc Interpretability: Unjustified Counterfactual Explanationscited by 97
- 2022AIM: An Adaptive and Iterative Mechanism for Differentially Private Synthetic Datacited by 88
- 2021NICE: An Algorithm for Nearest Instance Counterfactual Explanationscited by 62
- 2024Comprehensive Exploration of Synthetic Data Generation: A Surveycited by 59
- 2022TAPAS: a Toolbox for Adversarial Privacy Auditing of Synthetic Datacited by 56
8 more
- 2020Model extraction from counterfactual explanationscited by 46
- 2022Private Synthetic Data for Multitask Learning and Marginal Queriescited by 35
- 2022DualCF: Efficient Model Extraction Attack from Counterfactual Explanationscited by 31
- 2022The privacy issue of counterfactual explanations: explanation linkage attackscited by 20
- 2024Data Plagiarism Index: Characterizing the Privacy Risk of Data-Copying in Tabular Generative Modelscited by 12
- 2021TableGAN-MCA: Evaluating Membership Collisions of GAN-Synthesized Tabular Data Releasingcited by 7
- 2023Accurate, Explainable, and Private Models: Providing Recourse While Minimizing Training Data Leakagecited by 5
- 2025Privacy Auditing Synthetic Data Release through Local Likelihood Attackscited by 5
Built on by
Not cited yet.