Overfitting, robustness, and malicious algorithms: A study of potential causes of privacy risk in machine learning
Open accessThe DOI matches its Crossref record; 12 of the 12 references it cites were found in a registry or an index.
Abstract
Machine learning algorithms, when applied to sensitive data, pose a distinct threat to privacy. A growing body of prior work demonstrates that models produced by these algorithms may leak specific private information in the training data to an attacker, either through the models’ structure or their observable behavior. This article examines the factors that can allow a training set membership inference attacker or an attribute inference attacker to learn such information. Using both formal and empirical analyses, we illustrate a clear relationship between these factors and the privacy risk that arises in several popular machine learning algorithms. We find that overfitting is sufficient to allow an attacker to perform membership inference and, when the target attribute meets certain conditions about its influence, attribute inference attacks. We also explore the connection between membership inference and attribute inference, showing that there are deep connections between the two that lead to effective new attacks. We show that overfitting is not necessary for these attacks, demonstrating that other factors, such as robustness to norm-bounded input perturbations and malicious training algorithms, can also significantly increase the privacy risk. Notably, as robustness is intended to be a defense against attacks on the integrity of model predictions, these results suggest it may be difficult in some cases to simultaneously defend against privacy and integrity attacks.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Membership Inference Attacks Against Machine Learning Models
Shokri, R., Stronati, M., Song, C., & Shmatikov, V. (2017) · 2017 IEEE Symposium on Security and Privacy (SP) · doi:10.1109/sp.2017.41 · Cited by 4,754 · cited 10×
“This is slightly different from the sort of membership inference described in some prior work [39,54], which distinguishes between z ∼ S and z ∼ D \ S .”
BackgroundMethodResultInfluentialThe Limitations of Deep Learning in Adversarial Settings
Papernot, N., McDaniel, P., Jha, S., Fredrikson, M., Celik, Z. B., & Swami, A. (2016) · 2016 IEEE European Symposium on Security and Privacy (EuroS&P) · doi:10.1109/eurosp.2016.36 · Cited by 3,993
Self-citationPractical Black-Box Attacks against Machine Learning
Papernot, N., McDaniel, P., Goodfellow, I., Jha, S., Celik, Z. B., & Swami, A. (2017) · Proceedings of the 2017 ACM on Asia Conference on Computer and Communications Security · doi:10.1145/3052973.3053009 · Cited by 3,495 · cited 2×
“Prior work on these attacks [26,46,47,59] has shown that it is often possible to make small, visually imperceptible changes to image features such that a deep convolutional neural network (CNN) classifier is “tricked” into labeling the resulting image as something completely different from the true…”
BackgroundSelf-citationTheoretically principled trade-off between robustness and accuracy
Zhang, H., Yu, Y., Jiao, J., Xing, E. P., Ghaoui, L. E., & Jordan, M. I. (2019) · arXiv · doi:10.48550/arxiv.1901.08573 · Cited by 1,560
Homer, N., Szelinger, S., Redman, M., Duggan, D., Tembe, W., Muehling, J., Pearson, J. V., Stephan, D. A., Nelson, S. F., & Craig, D. W. (2008) · PLoS Genetics · doi:10.1371/journal.pgen.1000167 · Cited by 1,221 · cited 2×
“A number of previous studies [21,27,28,50,55,62] have looked into membership attacks from statistics commonly published in genome-wide association studies (GWAS).”
BackgroundPrivacy Risk in Machine Learning: Analyzing the Connection to Overfitting
Yeom, S., Giacomelli, I., Fredrikson, M., & Jha, S. (2018) · 2018 IEEE 31st Computer Security Foundations Symposium (CSF) · doi:10.1109/csf.2018.00027 · Cited by 1,001
Self-citationMachine Learning Models that Remember Too Much
Song, C., Ristenpart, T., & Shmatikov, V. (2017) · Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/3133956.3134077 · Cited by 526 · cited 2×
“The recent work of Song et al. [57] considers a similar setting, wherein a malicious machine learning provider supplies a closed-source training algorithm to users with private data.”
BackgroundA Systematic Review of Re-Identification Attacks on Health Data
El Emam, K., Jonker, E., Arbuckle, L., & Malin, B. (2011) · PLoS ONE · doi:10.1371/journal.pone.0028071 · Cited by 427 · cited 1×
“A number of previous studies [21,27,28,50,55,62] have looked into membership attacks from statistics commonly published in genome-wide association studies (GWAS).”
Background"You Might Also Like:" Privacy Risks of Collaborative Filtering
Calandrino, J. A., Kilzer, A., Narayanan, A., Felten, E. W., & Shmatikov, V. (2011) · 2011 IEEE Symposium on Security and Privacy · doi:10.1109/sp.2011.40 · Cited by 311 · cited 1×
“[10] showed that temporal changes in recommendations given by collaborative filtering methods can reveal the inputs that caused those changes.”
BackgroundPrivacy Risks of Securing Machine Learning Models against Adversarial Examples
Song, L., Shokri, R., & Mittal, P. (2019) · Proceedings of the 2019 ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/3319535.3354211 · Cited by 229 · cited 1×
“[58] evaluate several different attacks that seek to extract membership information from robust models, showing that robustness can make a model more vulnerable to membership inference.”
Background
Cited by
Local and Central Differential Privacy for Robustness and Privacy in Federated Learning
Naseri, M., Hayes, J., & De Cristofaro, E. (2022) · Proceedings 2022 Network and Distributed System Security Symposium · doi:10.14722/ndss.2022.23054 · Cited by 188
Survey: Leakage and Privacy at Inference Time
Jegorova, M., Kaul, C., Mayor, C., O'Neil, A. Q., Weir, A., Murray-Smith, R., & Tsaftaris, S. A. (2022) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2022.3229593 · Cited by 77
Threats and Defenses in the Federated Learning Life Cycle: A Comprehensive Survey and Challenges
Li, Y., Guo, Z., Yang, N., Chen, H., Yuan, D., & Ding, W. (2025) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2025.3563537 · Cited by 38
“Second, model overfitting has been identified as a primary factor in the success of MIA [66], and methods for mitigating overfitting can potentially defend against such attacks.”
Metadata
- Authors
- Yeom, S., Giacomelli, I., Menaged, A., Fredrikson, M., & Jha, S.
- Venue
- Journal of Computer Security
- Publisher
- SAGE Publications
- Published
- 2019-10-22
- Volume
- 28
- Issue
- 1
- Pages
- 35-70
- ISSN
- 0926-227X (ISSN-L)
- 1875-8924
- OpenAlex
- W2981338466
- Cited by
- 39
- References
- 12
- FWCI
- 1.96
- Type
- journal article
- Language
- en
- Open access
- hybrid
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Overfitting, robustness, and malicious algorithms: A study of potential causes of privacy risk in machine learning. https://app.cochat.ai/papers/10.3233/jcs-191362
- The paper
- Yeom, S., Giacomelli, I., Menaged, A., Fredrikson, M., & Jha, S. (2020). Overfitting, robustness, and malicious algorithms: A study of potential causes of privacy risk in machine learning. Journal of Computer Security. https://doi.org/10.3233/jcs-191362
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.