arXiv subjects

Neal Mangaokar

Publications and source records attributed to Neal Mangaokar.

15 recordsLinked to original sources

arXiv2026

Inadvertent Context Leakage in Language Models

For AI agents to be useful beyond simple chat, they must hold sensitive user context such as calendars, credentials, health records, and financial data. We study whether the mere presence of such secrets in a model's context window introduces hidden correlations into the model's benign outputs, allowing reconstruction even when the model correctly refuses direct extraction. We further study whether an adversary can actively engineer prompts that amplify this effect, using the model as a covert carrier to transmit secrets through seemingly innocuous text. In both cases, this limited leakage is exploited using a novel adaptive attack that assumes black-box access to the underlying model. In controlled experiments across eight proprietary models, we find that 2-digit in-context secrets are reconstructed with near-perfect accuracy and 4-digit secrets at 82\% exact match, all from outputs the model produces in response to ordinary, non-adversarial requests. We observe that more capable models leak more: stronger instruction-following amplifies sensitivity to in-context secrets, suggesting leakage is a byproduct of capability as opposed to a patchable bug. We show this leakage enables two practical attacks: (1) a trained classifier that infers semantic predicates about user memories (e.g., health conditions, financial events) from routine natural-language outputs, and (2) an RL-trained adversary that extracts full Social Security Numbers from a production-style agent.

cs.LG↗

arXiv2026

Muse Spark Safety & Preparedness Report

Muse Spark is the latest large language model developed by Meta. In this report, we first present evaluations for catastrophic risk domains under Meta's Advanced AI Scaling Framework, along with the evidence that informed our launch decision. We then discuss additional considerations, such as Muse Spark's broader content safety and behavioral profile, that are relevant to overall safety but fall outside the catastrophic risk domains governed by the Framework. Our preparedness results covering Chemical and Biological, Cybersecurity, and Loss of Control risks assess Muse Spark's deployment within Meta AI as presenting acceptable levels of residual risks under our Advanced AI Scaling Framework. We conducted a broad set of evaluations targeting dual-use and high-risk capabilities across these catastrophic risk domains. Those evaluations identified elevated risks prior to mitigations, with Chemical and Biological capabilities assessed as likely reaching the "high risk" category under the Advanced AI Scaling Framework before safeguards were applied. We have implemented a multi-layered set of mitigations that address the identified risks, and Muse Spark demonstrates state-of-the-art refusal across a range of benchmarks related to hazardous workflows in chemistry and biology. We therefore release Muse Spark as the underlying model of Meta AI.

cs.CY↗

arXiv2025

CIMemories: A Compositional Benchmark for Contextual Integrity of Persistent Memory in LLMs

Large Language Models (LLMs) increasingly use persistent memory from past interactions to enhance personalization and task performance. However, this memory introduces critical risks when sensitive information is revealed in inappropriate contexts. We present CIMemories, a benchmark for evaluating whether LLMs appropriately control information flow from memory based on task context. CIMemories uses synthetic user profiles with over 100 attributes per user, paired with diverse task contexts in which each attribute may be essential for some tasks but inappropriate for others. Our evaluation reveals that frontier models exhibit up to 69% attribute-level violations (leaking information inappropriately), with lower violation rates often coming at the cost of task utility. Violations accumulate across both tasks and runs: as usage increases from 1 to 40 tasks, GPT-5's violations rise from 0.1% to 9.6%, reaching 25.1% when the same prompt is executed 5 times, revealing arbitrary and unstable behavior in which models leak different attributes for identical prompts. Privacy-conscious prompting does not solve this - models overgeneralize, sharing everything or nothing rather than making nuanced, context-dependent decisions. These findings reveal fundamental limitations that require contextually aware reasoning capabilities, not just better prompting or scaling.

cs.CR↗

arXiv2025

Taming Data Challenges in ML-based Security Tasks Using Generative AI

Machine learning-based supervised classifiers are widely used for security tasks, and their improvement has been largely focused on algorithmic advancements. We argue that data challenges that negatively impact the performance of these classifiers have received limited attention. We address the following research question: Can developments in Generative AI (GenAI) address these data challenges and improve classifier performance? We propose augmenting training datasets with synthetic data generated using GenAI techniques to improve classifier generalization. We evaluate this approach across 7 diverse security tasks using 6 state-of-the-art GenAI methods and introduce a novel GenAI scheme called Nimai that enables highly controlled data synthesis. We find that GenAI techniques can significantly improve the performance of security classifiers, achieving improvements of up to 32.6% even in severely data-constrained settings (only ~180 training samples). Furthermore, we demonstrate that GenAI can facilitate rapid adaptation to concept drift post-deployment, requiring minimal labeling in the adjustment process. Despite successes, our study finds that some GenAI schemes struggle to initialize (train and produce data) on certain security tasks. We also identify characteristics of specific tasks, such as noisy labels, overlapping class distributions, and sparse feature vectors, which hinder performance boost using GenAI. We believe that our study will drive the development of future GenAI tools designed for security tasks.

cs.CR↗

arXiv2025

What Really is a Member? Discrediting Membership Inference via Poisoning

Membership inference tests aim to determine whether a particular data point was included in a language model's training set. However, recent works have shown that such tests often fail under the strict definition of membership based on exact matching, and have suggested relaxing this definition to include semantic neighbors as members as well. In this work, we show that membership inference tests are still unreliable under this relaxation - it is possible to poison the training dataset in a way that causes the test to produce incorrect predictions for a target point. We theoretically reveal a trade-off between a test's accuracy and its robustness to poisoning. We also present a concrete instantiation of this poisoning attack and empirically validate its effectiveness. Our results show that it can degrade the performance of existing tests to well below random.

cs.LG↗

arXiv2025

Harmful Terms and Where to Find Them: Measuring and Modeling Unfavorable Financial Terms and Conditions in Shopping Websites at Scale

Terms and conditions for online shopping websites often contain terms that can have significant financial consequences for customers. Despite their impact, there is currently no comprehensive understanding of the types and potential risks associated with unfavorable financial terms. Furthermore, there are no publicly available detection systems or datasets to systematically identify or mitigate these terms. In this paper, we take the first steps toward solving this problem with three key contributions. \textit{First}, we introduce \textit{TermMiner}, an automated data collection and topic modeling pipeline to understand the landscape of unfavorable financial terms. \textit{Second}, we create \textit{ShopTC-100K}, a dataset of terms and conditions from shopping websites in the Tranco top 100K list, comprising 1.8 million terms from 8,251 websites. Consequently, we develop a taxonomy of 22 types from 4 categories of unfavorable financial terms -- spanning purchase, post-purchase, account termination, and legal aspects. \textit{Third}, we build \textit{TermLens}, an automated detector that uses Large Language Models (LLMs) to identify unfavorable financial terms. Fine-tuned on an annotated dataset, \textit{TermLens} achieves an F1 score of 94.6\% and a false positive rate of 2.3\% using GPT-4o. When applied to shopping websites from the Tranco top 100K, we find that 42.06\% of these sites contain at least one unfavorable financial term, with such terms being more prevalent on less popular websites. Case studies further highlight the financial risks and customer dissatisfaction associated with unfavorable financial terms, as well as the limitations of existing ecosystem defenses.

cs.CR↗

arXiv2024

PRP: Propagating Universal Perturbations to Attack Large Language Model Guard-Rails

Large language models (LLMs) are typically aligned to be harmless to humans. Unfortunately, recent work has shown that such models are susceptible to automated jailbreak attacks that induce them to generate harmful content. More recent LLMs often incorporate an additional layer of defense, a Guard Model, which is a second LLM that is designed to check and moderate the output response of the primary LLM. Our key contribution is to show a novel attack strategy, PRP, that is successful against several open-source (e.g., Llama 2) and closed-source (e.g., GPT 3.5) implementations of Guard Models. PRP leverages a two step prefix-based attack that operates by (a) constructing a universal adversarial prefix for the Guard Model, and (b) propagating this prefix to the response. We find that this procedure is effective across multiple threat models, including ones in which the adversary has no access to the Guard Model at all. Our work suggests that further advances are required on defenses and Guard Models before they can be considered effective.

cs.CR↗

arXiv2023

Theoretically Principled Trade-off for Stateful Defenses against Query-Based Black-Box Attacks

Adversarial examples threaten the integrity of machine learning systems with alarming success rates even under constrained black-box conditions. Stateful defenses have emerged as an effective countermeasure, detecting potential attacks by maintaining a buffer of recent queries and detecting new queries that are too similar. However, these defenses fundamentally pose a trade-off between attack detection and false positive rates, and this trade-off is typically optimized by hand-picking feature extractors and similarity thresholds that empirically work well. There is little current understanding as to the formal limits of this trade-off and the exact properties of the feature extractors/underlying problem domain that influence it. This work aims to address this gap by offering a theoretical characterization of the trade-off between detection and false positive rates for stateful defenses. We provide upper bounds for detection rates of a general class of feature extractors and analyze the impact of this trade-off on the convergence of black-box attacks. We then support our theoretical findings with empirical evaluations across multiple datasets and stateful defenses.

cs.LG↗

arXiv2023

Stateful Defenses for Machine Learning Models Are Not Yet Secure Against Black-box Attacks

Recent work has proposed stateful defense models (SDMs) as a compelling strategy to defend against a black-box attacker who only has query access to the model, as is common for online machine learning platforms. Such stateful defenses aim to defend against black-box attacks by tracking the query history and detecting and rejecting queries that are "similar" and thus preventing black-box attacks from finding useful gradients and making progress towards finding adversarial attacks within a reasonable query budget. Recent SDMs (e.g., Blacklight and PIHA) have shown remarkable success in defending against state-of-the-art black-box attacks. In this paper, we show that SDMs are highly vulnerable to a new class of adaptive black-box attacks. We propose a novel adaptive black-box attack strategy called Oracle-guided Adaptive Rejection Sampling (OARS) that involves two stages: (1) use initial query patterns to infer key properties about an SDM's defense; and, (2) leverage those extracted properties to design subsequent query patterns to evade the SDM's defense while making progress towards finding adversarial inputs. OARS is broadly applicable as an enhancement to existing black-box attacks - we show how to apply the strategy to enhance six common black-box attacks to be more effective against current class of SDMs. For example, OARS-enhanced versions of black-box attacks improved attack success rate against recent stateful defenses from almost 0% to to almost 100% for multiple datasets within reasonable query budgets.

cs.CR↗

arXiv2022

Defending Object Detectors against Patch Attacks with Out-of-Distribution Smoothing

Patch attacks against object detectors have been of recent interest due to their being physically realizable and more closely aligned with practical systems. In response to this threat, many new defenses have been proposed that train a patch segmenter model to detect and remove the patch before the image is passed to the downstream model. We unify these approaches with a flexible framework, OODSmoother, which characterizes the properties of approaches that aim to remove adversarial patches. This framework naturally guides us to design 1) a novel adaptive attack that breaks existing patch attack defenses on object detectors, and 2) a novel defense approach SemPrior that takes advantage of semantic priors. Our key insight behind SemPrior is that the existing machine learning-based patch detectors struggle to learn semantic priors and that explicitly incorporating them can improve performance. We find that SemPrior alone provides up to a 40% gain, or up to a 60% gain when combined with existing defenses.

cs.LG↗

arXiv2022

D4: Detection of Adversarial Diffusion Deepfakes Using Disjoint Ensembles

Detecting diffusion-generated deepfake images remains an open problem. Current detection methods fail against an adversary who adds imperceptible adversarial perturbations to the deepfake to evade detection. In this work, we propose Disjoint Diffusion Deepfake Detection (D4), a deepfake detector designed to improve black-box adversarial robustness beyond de facto solutions such as adversarial training. D4 uses an ensemble of models over disjoint subsets of the frequency spectrum to significantly improve adversarial robustness. Our key insight is to leverage a redundancy in the frequency domain and apply a saliency partitioning technique to disjointly distribute frequency components across multiple models. We formally prove that these disjoint ensembles lead to a reduction in the dimensionality of the input subspace where adversarial deepfakes lie, thereby making adversarial deepfakes harder to find for black-box attacks. We then empirically validate the D4 method against several black-box attacks and find that D4 significantly outperforms existing state-of-the-art defenses applied to diffusion-generated deepfake detection. We also demonstrate that D4 provides robustness against adversarial deepfakes from unseen data distributions as well as unseen generative techniques.

cs.LG↗

arXiv2021

Jekyll: Attacking Medical Image Diagnostics using Deep Generative Models

Advances in deep neural networks (DNNs) have shown tremendous promise in the medical domain. However, the deep learning tools that are helping the domain, can also be used against it. Given the prevalence of fraud in the healthcare domain, it is important to consider the adversarial use of DNNs in manipulating sensitive data that is crucial to patient healthcare. In this work, we present the design and implementation of a DNN-based image translation attack on biomedical imagery. More specifically, we propose Jekyll, a neural style transfer framework that takes as input a biomedical image of a patient and translates it to a new image that indicates an attacker-chosen disease condition. The potential for fraudulent claims based on such generated 'fake' medical images is significant, and we demonstrate successful attacks on both X-rays and retinal fundus image modalities. We show that these attacks manage to mislead both medical professionals and algorithmic detection schemes. Lastly, we also investigate defensive measures based on machine learning to detect images generated by Jekyll.

cs.CR↗

arXiv2021

Deepfake Videos in the Wild: Analysis and Detection

AI-manipulated videos, commonly known as deepfakes, are an emerging problem. Recently, researchers in academia and industry have contributed several (self-created) benchmark deepfake datasets, and deepfake detection algorithms. However, little effort has gone towards understanding deepfake videos in the wild, leading to a limited understanding of the real-world applicability of research contributions in this space. Even if detection schemes are shown to perform well on existing datasets, it is unclear how well the methods generalize to real-world deepfakes. To bridge this gap in knowledge, we make the following contributions: First, we collect and present the largest dataset of deepfake videos in the wild, containing 1,869 videos from YouTube and Bilibili, and extract over 4.8M frames of content. Second, we present a comprehensive analysis of the growth patterns, popularity, creators, manipulation strategies, and production methods of deepfake content in the real-world. Third, we systematically evaluate existing defenses using our new dataset, and observe that they are not ready for deployment in the real-world. Fourth, we explore the potential for transfer learning schemes and competition-winning techniques to improve defenses.

cs.CR↗

arXiv2021

T-Miner: A Generative Approach to Defend Against Trojan Attacks on DNN-based Text Classification

Deep Neural Network (DNN) classifiers are known to be vulnerable to Trojan or backdoor attacks, where the classifier is manipulated such that it misclassifies any input containing an attacker-determined Trojan trigger. Backdoors compromise a model's integrity, thereby posing a severe threat to the landscape of DNN-based classification. While multiple defenses against such attacks exist for classifiers in the image domain, there have been limited efforts to protect classifiers in the text domain. We present Trojan-Miner (T-Miner) -- a defense framework for Trojan attacks on DNN-based text classifiers. T-Miner employs a sequence-to-sequence (seq-2-seq) generative model that probes the suspicious classifier and learns to produce text sequences that are likely to contain the Trojan trigger. T-Miner then analyzes the text produced by the generative model to determine if they contain trigger phrases, and correspondingly, whether the tested classifier has a backdoor. T-Miner requires no access to the training dataset or clean inputs of the suspicious classifier, and instead uses synthetically crafted "nonsensical" text inputs to train the generative model. We extensively evaluate T-Miner on 1100 model instances spanning 3 ubiquitous DNN model architectures, 5 different classification tasks, and a variety of trigger phrases. We show that T-Miner detects Trojan and clean models with a 98.75% overall accuracy, while achieving low false positives on clean models. We also show that T-Miner is robust against a variety of targeted, advanced attacks from an adaptive attacker.

cs.CR↗

arXiv2020

GRAPHITE: Generating Automatic Physical Examples for Machine-Learning Attacks on Computer Vision Systems

This paper investigates an adversary's ease of attack in generating adversarial examples for real-world scenarios. We address three key requirements for practical attacks for the real-world: 1) automatically constraining the size and shape of the attack so it can be applied with stickers, 2) transform-robustness, i.e., robustness of a attack to environmental physical variations such as viewpoint and lighting changes, and 3) supporting attacks in not only white-box, but also black-box hard-label scenarios, so that the adversary can attack proprietary models. In this work, we propose GRAPHITE, an efficient and general framework for generating attacks that satisfy the above three key requirements. GRAPHITE takes advantage of transform-robustness, a metric based on expectation over transforms (EoT), to automatically generate small masks and optimize with gradient-free optimization. GRAPHITE is also flexible as it can easily trade-off transform-robustness, perturbation size, and query count in black-box settings. On a GTSRB model in a hard-label black-box setting, we are able to find attacks on all possible 1,806 victim-target class pairs with averages of 77.8% transform-robustness, perturbation size of 16.63% of the victim images, and 126K queries per pair. For digital-only attacks where achieving transform-robustness is not a requirement, GRAPHITE is able to find successful small-patch attacks with an average of only 566 queries for 92.2% of victim-target pairs. GRAPHITE is also able to find successful attacks using perturbations that modify small areas of the input image against PatchGuard, a recently proposed defense against patch-based attacks.

cs.CR↗