Training data and privacy
Leakage of personal or proprietary data through training, memorization, inference attacks or careless logging.
- All items
- 32
- Last 90 days
- 14
- Change
- +133%vs 6 before
Items per month
| Month | Items |
|---|---|
| Jun 2025 | 0 |
| Jul 2025 | 0 |
| Aug 2025 | 0 |
| Sep 2025 | 1 |
| Oct 2025 | 2 |
| Nov 2025 | 0 |
| Dec 2025 | 0 |
| Jan 2026 | 1 |
| Feb 2026 | 1 |
| Mar 2026 | 2 |
| Apr 2026 | 2 |
| May 2026 | 2 |
| Jun 2026 | 2 |
| Jul 2026 | 2 |
| Aug 2026 | 3 |
| Sep 2026 | 0 |
| Oct 2026 | 8 |
| Nov 2026 | 5 |
32 items
<em>Infer-Shield</em>: Defending against membership inference attacks in heterogeneous federated learning via adaptive distillation
May 25, 2026InfoResearchPeer-reviewedSecurityResearchElsevier Security JournalsLearning to Defend: Auto-Augmentation Search Against Model Inversion Attacks
May 6, 2026InfoResearchPeer-reviewedSecurityPrivacyResearchThis research paper introduces DAAS (Defense via Auto-Augmentation Search) against Model Inversion Attacks, which can recover private training data from model weights or outputs. The authors note that existing defenses offer incomplete protection and that input-level defenses have been limited to simple transformations. DAAS automatically identifies augmentation candidates with strong privacy-utility trade-offs from a large pool, and the authors report superior defense performance across models, datasets and attacks.
IEEE Xplore (Security & AI Journals)GHSA-4jpm-cgx2-8h37: Flowise: Sensitive Data Leak in public-chatbotConfig
Apr 16, 2026HighVulnerabilitySecurityPrivacyThe /api/v1/public-chatbotConfig/:id endpoint in Flowise returns the full flowData object without authentication or sanitization, through the getSinglePublicChatbotConfig function in packages/server/src/services/chatflows/index.ts. An attacker who knows only a chatflow UUID, which can be obtained from embedded chat widgets, referrer headers or logs, can retrieve plaintext password-type fields such as unstructuredAPIKey and HTTP Authorization headers. The impact covers all Flowise Cloud users with such chatflows and self-hosted instances exposed to the internet.
GitHub Advisory DatabaseDiffMI: Breaking Face Recognition Privacy via Diffusion-Driven Training-Free Model Inversion
Apr 16, 2026LowResearchPeer-reviewedSecurityPrivacyResearchDiffMI is a diffusion-driven, training-free model inversion attack that recovers identity information from face recognition systems that map facial images to embeddings. The method combines latent code initialization, ranked adversarial refinement, and a confidence-aware optimization objective, and it applies directly to unseen target identities. The authors report 84.42%–92.87% attack success rates against inversion-resilient systems, 4.01%–9.82% higher than the best prior training-free GAN-based approach. The implementation is available at https://github.com/azrealwang/DiffMI.
IEEE Xplore (Security & AI Journals)Meta AI agent’s instruction causes large sensitive data leak to employees
Mar 20, 2026LowNewsSecurityPrivacyAn AI agent at Meta gave an engineer a solution to an engineering question posted on an internal forum, and the engineer implemented it. The resulting exposure made a large amount of sensitive user and company data visible to engineers for two hours, which Meta confirmed.
The Guardian TechnologyExtracting Training Dialogue Data From Large Language Model-Based Task Bots
Mar 5, 2026LowResearchPeer-reviewedSecurityPrivacyResearchResearchers study whether LLM-based task-oriented dialogue systems memorize their training dialogue data, which can include phone numbers and full travel schedules. They evaluate existing training data extraction attacks, find that task-oriented dialogue traits make them ineffective, and propose new attack techniques for response sampling and membership inference. Their method extracts thousands of training dialogue-state labels with best-case precision above 70%, and the paper analyzes the factors that influence this memorization.
IEEE Xplore (Security & AI Journals)Model Inversion Attack Against Federated Unlearning
Feb 19, 2026InfoResearchPeer-reviewedSecurityPrivacyThis paper presents FUIA, a federated unlearning inversion attack that exposes privacy leakage in federated unlearning (FU), the method used to remove specific data's influence from federated learning models. The first systematic study of FU privacy vulnerabilities, it applies to sample, client and class unlearning. An honest-but-curious server records parameter changes during unlearning, infers gradients of forgotten data, and reconstructs its features or labels across multiple benchmark datasets and FU methods.
Fix: The paper explores two potential defense strategies that trade off privacy protection against model performance. No specific defense details are given in the source text.
IEEE Xplore (Security & AI Journals)BlindU: Blind Machine Unlearning Without Revealing Erasing Data
Jan 15, 2026InfoResearchPeer-reviewedResearchPrivacyBlindU is a machine unlearning method for federated learning that removes a user's data contribution without uploading the erasing data to the server. The user locally produces compressed, privacy-preserving representations through an information bottleneck encoder, and the server unlearns using only those representations and their labels. The authors add a noise-free differential privacy masking step before compression and report better privacy protection and unlearning effectiveness than the best existing privacy-preserving benchmarks.
IEEE Xplore (Security & AI Journals)MaxDiv: Zero-Shot Machine Unlearning via Distributionally Divergent Erasing Samples
Oct 30, 2025InfoResearchPeer-reviewedResearchPrivacyMaxDiv is a zero-shot machine unlearning method that erases knowledge from a model without access to its original training data. It generates erasing samples by negating the distributions of the data to be forgotten, and adds knowledge distillation to prevent catastrophic forgetting of retained information. Evaluations on MNIST, SVHN, CIFAR-10, CIFAR-100 and TinyImageNet reportedly show superior performance over current state-of-the-art methods.
IEEE Xplore (Security & AI Journals)Really Unlearned? Verifying Machine Unlearning via Influential Sample Pairs
Oct 13, 2025InfoResearchPeer-reviewedResearchSecurityIndirectVerify is a formal scheme for verifying whether machine unlearning requests have actually been executed, addressing a gap where existing verification relies on backdoor or membership inference attacks that a model provider can bypass through rapid fine-tuning. The authors generate influential sample pairs, trigger samples and reaction samples, using a perturbation-based method that alters only a small fraction of training samples so that reaction samples are misclassified. The paper claims this indirect influence provides greater robustness against bypassing than schemes that reuse the same samples throughout.
IEEE Xplore (Security & AI Journals)SMS: Self-Supervised Model Seeding for Verification of Machine Unlearning
Sep 29, 2025InfoResearchPeer-reviewedSecurityPrivacyResearchThe paper proposes Self-supervised Model Seeding (SMS), a scheme for verifying that a model has unlearned genuine user samples rather than only backdoored ones. SMS links user-specific seeds, original samples and the model through a self-supervised seeding task trained jointly with the primary service task. The authors report effectiveness across three datasets, several model architectures, and exact and approximate unlearning benchmarks.
IEEE Xplore (Security & AI Journals)CVE-2024-5206: scikit-learn TfidfVectorizer leaks sensitive training tokens via stop_words_
Jun 6, 2024MediumVulnerabilitySecurityPrivacyCVE-2024-5206CVE-2024-5206 affects scikit-learn's TfidfVectorizer in versions up to and including 1.4.1.post1. The vectorizer stores every token from the training data in its `stop_words_` attribute, not only the tokens TF-IDF needs, so tokens meant to be discarded, such as passwords or keys, can be exposed. The impact depends on the data the vectorizer processes.
Fix: Fixed in 1.5.0
NVD/CVE Database
Topic added 2026-10-09. An item belongs to this topic when its title matches one of the topic's patterns or its summary mentions the topic at least twice. Report a wrong match with the feedback button on the item.