arXiv subjects

Ninghui Li

Publications and source records attributed to Ninghui Li.

At least 37 records · Page 2Linked to original sources

arXiv2025

LLM Agents Should Employ Security Principles

Large Language Model (LLM) agents show considerable promise for automating complex tasks using contextual reasoning; however, interactions involving multiple agents and the system's susceptibility to prompt injection and other forms of context manipulation introduce new vulnerabilities related to privacy leakage and system exploitation. This position paper argues that the well-established design principles in information security, which are commonly referred to as security principles, should be employed when deploying LLM agents at scale. Design principles such as defense-in-depth, least privilege, complete mediation, and psychological acceptability have helped guide the design of mechanisms for securing information systems over the last five decades, and we argue that their explicit and conscientious adoption will help secure agentic systems. To illustrate this approach, we introduce AgentSandbox, a conceptual framework embedding these security principles to provide safeguards throughout an agent's life-cycle. We evaluate with state-of-the-art LLMs along three dimensions: benign utility, attack utility, and attack success rate. AgentSandbox maintains high utility for its intended functions under both benign and adversarial evaluations while substantially mitigating privacy risks. By embedding secure design principles as foundational elements within emerging LLM agent protocols, we aim to promote trustworthy agent ecosystems aligned with user privacy expectations and evolving regulatory requirements.

cs.CR↗

arXiv2025

The Federation Strikes Back: A Survey of Federated Learning Privacy Attacks, Defenses, Applications, and Policy Landscape

Deep learning has shown incredible potential across a wide array of tasks, and accompanied by this growth has been an insatiable appetite for data. However, a large amount of data needed for enabling deep learning is stored on personal devices, and recent concerns on privacy have further highlighted challenges for accessing such data. As a result, federated learning (FL) has emerged as an important privacy-preserving technology that enables collaborative training of machine learning models without the need to send the raw, potentially sensitive, data to a central server. However, the fundamental premise that sending model updates to a server is privacy-preserving only holds if the updates cannot be "reverse engineered" to infer information about the private training data. It has been shown under a wide variety of settings that this privacy premise does not hold. In this survey paper, we provide a comprehensive literature review of the different privacy attacks and defense methods in FL. We identify the current limitations of these attacks and highlight the settings in which the privacy of an FL client can be broken. We further dissect some of the successful industry applications of FL and draw lessons for future successful adoption. We survey the emerging landscape of privacy regulation for FL and conclude with future directions for taking FL toward the cherished goal of generating accurate models while preserving the privacy of the data from its participants.

cs.CR↗

arXiv2025

CENSOR: Defense Against Gradient Inversion via Orthogonal Subspace Bayesian Sampling

Federated learning collaboratively trains a neural network on a global server, where each local client receives the current global model weights and sends back parameter updates (gradients) based on its local private data. The process of sending these model updates may leak client's private data information. Existing gradient inversion attacks can exploit this vulnerability to recover private training instances from a client's gradient vectors. Recently, researchers have proposed advanced gradient inversion techniques that existing defenses struggle to handle effectively. In this work, we present a novel defense tailored for large neural network models. Our defense capitalizes on the high dimensionality of the model parameters to perturb gradients within a subspace orthogonal to the original gradient. By leveraging cold posteriors over orthogonal subspaces, our defense implements a refined gradient update mechanism. This enables the selection of an optimal gradient that not only safeguards against gradient inversion attacks but also maintains model utility. We conduct comprehensive experiments across three different datasets and evaluate our defense against various state-of-the-art attacks and defenses. Code is available at https://censor-gradient.github.io.

cs.LG↗

arXiv2025

On the Robustness of LDP Protocols for Numerical Attributes under Data Poisoning Attacks

Recent studies reveal that local differential privacy (LDP) protocols are vulnerable to data poisoning attacks where an attacker can manipulate the final estimate on the server by leveraging the characteristics of LDP and sending carefully crafted data from a small fraction of controlled local clients. This vulnerability raises concerns regarding the robustness and reliability of LDP in hostile environments. In this paper, we conduct a systematic investigation of the robustness of state-of-the-art LDP protocols for numerical attributes, i.e., categorical frequency oracles (CFOs) with binning and consistency, and distribution reconstruction. We evaluate protocol robustness through an attack-driven approach and propose new metrics for cross-protocol attack gain measurement. The results indicate that Square Wave and CFO-based protocols in the Server setting are more robust against the attack compared to the CFO-based protocols in the User setting. Our evaluation also unfolds new relationships between LDP security and its inherent design choices. We found that the hash domain size in local-hashing-based LDP has a profound impact on protocol robustness beyond the well-known effect on utility. Further, we propose a zero-shot attack detection by leveraging the rich reconstructed distribution information. The experiment show that our detection significantly improves the existing methods and effectively identifies data manipulation in challenging scenarios.

cs.CR↗

arXiv2024

MIST: Defending Against Membership Inference Attacks Through Membership-Invariant Subspace Training

In Member Inference (MI) attacks, the adversary try to determine whether an instance is used to train a machine learning (ML) model. MI attacks are a major privacy concern when using private data to train ML models. Most MI attacks in the literature take advantage of the fact that ML models are trained to fit the training data well, and thus have very low loss on training instances. Most defenses against MI attacks therefore try to make the model fit the training data less well. Doing so, however, generally results in lower accuracy. We observe that training instances have different degrees of vulnerability to MI attacks. Most instances will have low loss even when not included in training. For these instances, the model can fit them well without concerns of MI attacks. An effective defense only needs to (possibly implicitly) identify instances that are vulnerable to MI attacks and avoids overfitting them. A major challenge is how to achieve such an effect in an efficient training process. Leveraging two distinct recent advancements in representation learning: counterfactually-invariant representations and subspace learning methods, we introduce a novel Membership-Invariant Subspace Training (MIST) method to defend against MI attacks. MIST avoids overfitting the vulnerable instances without significant impact on other instances. We have conducted extensive experimental studies, comparing MIST with various other state-of-the-art (SOTA) MI defenses against several SOTA MI attacks. We find that MIST outperforms other defenses while resulting in minimal reduction in testing accuracy.

cs.CR↗

arXiv2023

Continuous Release of Data Streams under both Centralized and Local Differential Privacy

In this paper, we study the problem of publishing a stream of real-valued data satisfying differential privacy (DP). One major challenge is that the maximal possible value can be quite large; thus it is necessary to estimate a threshold so that numbers above it are truncated to reduce the amount of noise that is required to all the data. The estimation must be done based on the data in a private fashion. We develop such a method that uses the Exponential Mechanism with a quality function that approximates well the utility goal while maintaining a low sensitivity. Given the threshold, we then propose a novel online hierarchical method and several post-processing techniques. Building on these ideas, we formalize the steps into a framework for private publishing of stream data. Our framework consists of three components: a threshold optimizer that privately estimates the threshold, a perturber that adds calibrated noises to the stream, and a smoother that improves the result using post-processing. Within our framework, we design an algorithm satisfying the more stringent setting of DP called local DP (LDP). To our knowledge, this is the first LDP algorithm for publishing streaming data. Using four real-world datasets, we demonstrate that our mechanism outperforms the state-of-the-art by a factor of 6-10 orders of magnitude in terms of utility (measured by the mean squared error of answering a random range query).

cs.CR↗

arXiv2023

Differentially Private Vertical Federated Clustering

In many applications, multiple parties have private data regarding the same set of users but on disjoint sets of attributes, and a server wants to leverage the data to train a model. To enable model learning while protecting the privacy of the data subjects, we need vertical federated learning (VFL) techniques, where the data parties share only information for training the model, instead of the private data. However, it is challenging to ensure that the shared information maintains privacy while learning accurate models. To the best of our knowledge, the algorithm proposed in this paper is the first practical solution for differentially private vertical federated k-means clustering, where the server can obtain a set of global centers with a provable differential privacy guarantee. Our algorithm assumes an untrusted central server that aggregates differentially private local centers and membership encodings from local data parties. It builds a weighted grid as the synopsis of the global dataset based on the received information. Final centers are generated by running any k-means algorithm on the weighted grid. Our approach for grid weight estimation uses a novel, light-weight, and differentially private set intersection cardinality estimation algorithm based on the Flajolet-Martin sketch. To improve the estimation accuracy in the setting with more than two data parties, we further propose a refined version of the weights estimation algorithm and a parameter tuning strategy to reduce the final k-means utility to be close to that in the central private setting. We provide theoretical utility analysis and experimental evaluation results for the cluster centers computed by our algorithm and show that our approach performs better both theoretically and empirically than the two baselines based on existing techniques.

cs.CR↗

arXiv2023

Fine-grained Poisoning Attack to Local Differential Privacy Protocols for Mean and Variance Estimation

Although local differential privacy (LDP) protects individual users' data from inference by an untrusted data curator, recent studies show that an attacker can launch a data poisoning attack from the user side to inject carefully-crafted bogus data into the LDP protocols in order to maximally skew the final estimate by the data curator. In this work, we further advance this knowledge by proposing a new fine-grained attack, which allows the attacker to fine-tune and simultaneously manipulate mean and variance estimations that are popular analytical tasks for many real-world applications. To accomplish this goal, the attack leverages the characteristics of LDP to inject fake data into the output domain of the local LDP instance. We call our attack the output poisoning attack (OPA). We observe a security-privacy consistency where a small privacy loss enhances the security of LDP, which contradicts the known security-privacy trade-off from prior work. We further study the consistency and reveal a more holistic view of the threat landscape of data poisoning attacks on LDP. We comprehensively evaluate our attack against a baseline attack that intuitively provides false input to LDP. The experimental results show that OPA outperforms the baseline on three real-world datasets. We also propose a novel defense method that can recover the result accuracy from polluted data collection and offer insight into the secure LDP design.

cs.CR↗

arXiv2023

PolyScope: Multi-Policy Access Control Analysis to Triage Android Scoped Storage

Android's filesystem access control is a crucial aspect of its system integrity. It utilizes a combination of mandatory access controls, such as SELinux, and discretionary access controls, like Unix permissions, along with specialized access controls such as Android permissions to safeguard OEM and Android services from third-party applications. However, when OEMs introduce differentiating features, they often create vulnerabilities due to their inability to properly reconfigure this complex policy combination. To address this, we introduce the POLYSCOPE tool, which triages Android filesystem access control policies to identify attack operations - authorized operations that may be exploited by adversaries to elevate their privileges. POLYSCOPE has three significant advantages over prior analyses: it allows for the independent extension and analysis of individual policy models, understands the flexibility untrusted parties have in modifying access control policies, and can identify attack operations that system configurations permit. We demonstrate the effectiveness of POLYSCOPE by examining the impact of Scoped Storage on Android, revealing that it reduces the number of attack operations possible on external storage resources by over 50%. However, because OEMs only partially adopt Scoped Storage, we also uncover two previously unknown vulnerabilities, demonstrating how POLYSCOPE can assess an ideal scenario where all apps comply with Scoped Storage, which can reduce the number of untrusted parties accessing attack operations by over 65% on OEM systems. POLYSCOPE thus helps Android OEMs evaluate complex access control policies to pinpoint the attack operations that require further examination.

cs.CR↗

arXiv2022

Using Illustrations to Communicate Differential Privacy Trust Models: An Investigation of Users' Comprehension, Perception, and Data Sharing Decision

Proper communication is key to the adoption and implementation of differential privacy (DP). However, a prior study found that laypeople did not understand the data perturbation processes of DP and how DP noise protects their sensitive personal information. Consequently, they distrusted the techniques and chose to opt out of participating. In this project, we designed explanative illustrations of three DP models (Central DP, Local DP, Shuffler DP) to help laypeople conceptualize how random noise is added to protect individuals' privacy and preserve group utility. Following pilot surveys and interview studies, we conducted two online experiments (N = 595) examining participants' comprehension, privacy and utility perception, and data-sharing decisions across the three DP models. Besides the comparisons across the three models, we varied the noise levels of each model. We found that the illustrations can be effective in communicating DP to the participants. Given an adequate comprehension of DP, participants preferred strong privacy protection for a certain type of data usage scenarios (i.e., commercial interests) at both the model level and the noise level. We also obtained empirical evidence showing participants' acceptance of the Shuffler DP model for data privacy protection. Our findings have implications for multiple stakeholders for user-centered deployments of differential privacy, including app developers, DP model developers, data curators, and online users.

cs.CR↗

arXiv2022

Are Your Sensitive Attributes Private? Novel Model Inversion Attribute Inference Attacks on Classification Models

Increasing use of machine learning (ML) technologies in privacy-sensitive domains such as medical diagnoses, lifestyle predictions, and business decisions highlights the need to better understand if these ML technologies are introducing leakage of sensitive and proprietary training data. In this paper, we focus on model inversion attacks where the adversary knows non-sensitive attributes about records in the training data and aims to infer the value of a sensitive attribute unknown to the adversary, using only black-box access to the target classification model. We first devise a novel confidence score-based model inversion attribute inference attack that significantly outperforms the state-of-the-art. We then introduce a label-only model inversion attack that relies only on the model's predicted labels but still matches our confidence score-based attack in terms of attack effectiveness. We also extend our attacks to the scenario where some of the other (non-sensitive) attributes of a target record are unknown to the adversary. We evaluate our attacks on two types of machine learning models, decision tree and deep neural network, trained on three real datasets. Moreover, we empirically demonstrate the disparate vulnerability of model inversion attacks, i.e., specific groups in the training dataset (grouped by gender, race, etc.) could be more vulnerable to model inversion attacks.

cs.CR↗

arXiv2021

DPSyn: Experiences in the NIST Differential Privacy Data Synthesis Challenges

We summarize the experience of participating in two differential privacy competitions organized by the National Institute of Standards and Technology (NIST). In this paper, we document our experiences in the competition, the approaches we have used, the lessons we have learned, and our call to the research community to further bridge the gap between theory and practice in DP research.

cs.CR↗

arXiv2021

Membership Inference Attacks and Defenses in Classification Models

We study the membership inference (MI) attack against classifiers, where the attacker's goal is to determine whether a data instance was used for training the classifier. Through systematic cataloging of existing MI attacks and extensive experimental evaluations of them, we find that a model's vulnerability to MI attacks is tightly related to the generalization gap -- the difference between training accuracy and test accuracy. We then propose a defense against MI attacks that aims to close the gap by intentionally reduces the training accuracy. More specifically, the training process attempts to match the training and validation accuracies, by means of a new {\em set regularizer} using the Maximum Mean Discrepancy between the softmax output empirical distributions of the training and validation sets. Our experimental results show that combining this approach with another simple defense (mix-up training) significantly improves state-of-the-art defense against MI attacks, with minimal impact on testing accuracy.

cs.CR↗

arXiv2020

PrivSyn: Differentially Private Data Synthesis

In differential privacy (DP), a challenging problem is to generate synthetic datasets that efficiently capture the useful information in the private data. The synthetic dataset enables any task to be done without privacy concern and modification to existing algorithms. In this paper, we present PrivSyn, the first automatic synthetic data generation method that can handle general tabular datasets (with 100 attributes and domain size $>2^{500}$). PrivSyn is composed of a new method to automatically and privately identify correlations in the data, and a novel method to generate sample data from a dense graphic model. We extensively evaluate different methods on multiple datasets to demonstrate the performance of our method.

cs.CR↗

arXiv2020

PURE: A Framework for Analyzing Proximity-based Contact Tracing Protocols

Many proximity-based tracing (PCT) protocols have been proposed and deployed to combat the spreading of COVID-19. In this paper, we take a systematic approach to analyze PCT protocols. We identify a list of desired properties of a contact tracing design from the four aspects of Privacy, Utility, Resiliency, and Efficiency (PURE). We also identify two main design choices for PCT protocols: what information patients report to the server, and which party performs the matching. These two choices determine most of the PURE properties and enable us to conduct a comprehensive analysis and comparison of the existing protocols.

cs.CY↗

arXiv2020

Black-box Model Inversion Attribute Inference Attacks on Classification Models

Increasing use of ML technologies in privacy-sensitive domains such as medical diagnoses, lifestyle predictions, and business decisions highlights the need to better understand if these ML technologies are introducing leakages of sensitive and proprietary training data. In this paper, we focus on one kind of model inversion attacks, where the adversary knows non-sensitive attributes about instances in the training data and aims to infer the value of a sensitive attribute unknown to the adversary, using oracle access to the target classification model. We devise two novel model inversion attribute inference attacks -- confidence modeling-based attack and confidence score-based attack, and also extend our attack to the case where some of the other (non-sensitive) attributes are unknown to the adversary. Furthermore, while previous work uses accuracy as the metric to evaluate the effectiveness of attribute inference attacks, we find that accuracy is not informative when the sensitive attribute distribution is unbalanced. We identify two metrics that are better for evaluating attribute inference attacks, namely G-mean and Matthews correlation coefficient (MCC). We evaluate our attacks on two types of machine learning models, decision tree and deep neural network, trained with two real datasets. Experimental results show that our newly proposed attacks significantly outperform the state-of-the-art attacks. Moreover, we empirically show that specific groups in the training dataset (grouped by attributes, e.g., gender, race) could be more vulnerable to model inversion attacks. We also demonstrate that our attacks' performances are not impacted significantly when some of the other (non-sensitive) attributes are also unknown to the adversary.

cs.CR↗

arXiv2020

Answering Multi-Dimensional Range Queries under Local Differential Privacy

In this paper, we tackle the problem of answering multi-dimensional range queries under local differential privacy. There are three key technical challenges: capturing the correlations among attributes, avoiding the curse of dimensionality, and dealing with the large domains of attributes. None of the existing approaches satisfactorily deals with all three challenges. Overcoming these three challenges, we first propose an approach called Two-Dimensional Grids (TDG). Its main idea is to carefully use binning to partition the two-dimensional (2-D) domains of all attribute pairs into 2-D grids that can answer all 2-D range queries and then estimate the answer of a higher dimensional range query from the answers of the associated 2-D range queries. However, in order to reduce errors due to noises, coarse granularities are needed for each attribute in 2-D grids, losing fine-grained distribution information for individual attributes. To correct this deficiency, we further propose Hybrid-Dimensional Grids (HDG), which also introduces 1-D grids to capture finer-grained information on distribution of each individual attribute and combines information from 1-D and 2-D grids to answer range queries. To make HDG consistently effective, we provide a guideline for properly choosing granularities of grids based on an analysis of how different sources of errors are impacted by these choices. Extensive experiments conducted on real and synthetic datasets show that HDG can give a significant improvement over the existing approaches.

cs.CR↗

arXiv2020

Improving Frequency Estimation under Local Differential Privacy

Local Differential Privacy protocols are stochastic protocols used in data aggregation when individual users do not trust the data aggregator with their private data. In such protocols there is a fundamental tradeoff between user privacy and aggregator utility. In the setting of frequency estimation, established bounds on this tradeoff are either nonquantitative, or far from what is known to be attainable. In this paper, we use information-theoretical methods to significantly improve established bounds. We also show that the new bounds are attainable for binary inputs. Furthermore, our methods lead to improved frequency estimators, which we experimentally show to outperform state-of-the-art methods.

cs.CR↗