Cryptography and Security (cs.CR)

Abstract: Smartwatches enable the efficient collection of health data that can be used for research and comprehensive analysis to improve the health of individuals. In addition to the analysis capabilities, ensuring privacy when handling health data is a critical concern as the collection and analysis of such data become pervasive. Since health data contains sensitive information, it should be handled with responsibility and is therefore often treated anonymously. However, also the data itself can be exploited to reveal information and break anonymity. We propose a novel similarity-based re-identification attack on time-series health data and thereby unveil a significant vulnerability. Despite privacy measures that remove identifying information, our attack demonstrates that a brief amount of various sensor data from a target individual is adequate to possibly identify them within a database of other samples, solely based on sensor-level similarities. In our example scenario, where data owners leverage health data from smartwatches, findings show that we are able to correctly link the target data in two out of three cases. User privacy is thus already inherently threatened by the data itself and even when removing personal information.

Abstract: In this paper, we establish anti-concentration inequalities for additive noise mechanisms which achieve $f$-differential privacy ($f$-DP), a notion of privacy phrased in terms of a tradeoff function (a.k.a. ROC curve) $f$ which limits the ability of an adversary to determine which individuals were in the database. We show that canonical noise distributions (CNDs), proposed by Awan and Vadhan (2023), match the anti-concentration bounds at half-integer values, indicating that their tail behavior is near-optimal. We also show that all CNDs are sub-exponential, regardless of the $f$-DP guarantee. In the case of log-concave CNDs, we show that they are the stochastically smallest noise compared to any other noise distributions with the same privacy guarantee. In terms of integer-valued noise, we propose a new notion of discrete CND and prove that a discrete CND always exists, can be constructed by rounding a continuous CND, and that the discrete CND is unique when designed for a statistic with sensitivity 1. We further show that the discrete CND at sensitivity 1 is stochastically smallest compared to other integer-valued noises. Our theoretical results shed light on the different types of privacy guarantees possible in the $f$-DP framework and can be incorporated in more complex mechanisms to optimize performance.

Abstract: IP blacklists are widely used to increase network security by preventing communications with peers that have been marked as malicious. There are several commercial offerings as well as several free-of-charge blacklists maintained by volunteers on the web. Despite their wide adoption, the effectiveness of the different IP blacklists in real-world scenarios is still not clear. In this paper, we conduct a large-scale network monitoring study which provides insightful findings regarding the effectiveness of blacklists. The results collected over several hundred thousand IP hosts belonging to three distinct large production networks highlight that blacklists are often tuned for precision, with the result that many malicious activities, such as scanning, are completely undetected. The proposed instrumentation approach to detect IP scanning and suspicious activities is implemented with home-grown and open-source software. Our tools enable the creation of blacklists without the security risks posed by the deployment of honeypots.

Abstract: To enhance the security of text CAPTCHAs, various methods have been employed, such as adding the interference lines on the text, randomly distorting the characters, and overlapping multiple characters. These methods partly increase the difficulty of automated segmentation and recognition attacks. However, facing the rapid development of the end-to-end breaking algorithms, their security has been greatly weakened. The diffusion model is a novel image generation model that can generate the text images with deep fusion of characters and background images. In this paper, an image-click CAPTCHA scheme called Diff-CAPTCHA is proposed based on denoising diffusion models. The background image and characters of the CAPTCHA are treated as a whole to guide the generation process of a diffusion model, thus weakening the character features available for machine learning, enhancing the diversity of character features in the CAPTCHA, and increasing the difficulty of breaking algorithms. To evaluate the security of Diff-CAPTCHA, this paper develops several attack methods, including end-to-end attacks based on Faster R-CNN and two-stage attacks, and Diff-CAPTCHA is compared with three baseline schemes, including commercial CAPTCHA scheme and security-enhanced CAPTCHA scheme based on style transfer. The experimental results show that diffusion models can effectively enhance CAPTCHA security while maintaining good usability in human testing.

Abstract: Deploying machine learning (ML) models in the wild is challenging as it suffers from distribution shifts, where the model trained on an original domain cannot generalize well to unforeseen diverse transfer domains. To address this challenge, several test-time adaptation (TTA) methods have been proposed to improve the generalization ability of the target pre-trained models under test data to cope with the shifted distribution. The success of TTA can be credited to the continuous fine-tuning of the target model according to the distributional hint from the test samples during test time. Despite being powerful, it also opens a new attack surface, i.e., test-time poisoning attacks, which are substantially different from previous poisoning attacks that occur during the training time of ML models (i.e., adversaries cannot intervene in the training process). In this paper, we perform the first test-time poisoning attack against four mainstream TTA methods, including TTT, DUA, TENT, and RPL. Concretely, we generate poisoned samples based on the surrogate models and feed them to the target TTA models. Experimental results show that the TTA methods are generally vulnerable to test-time poisoning attacks. For instance, the adversary can feed as few as 10 poisoned samples to degrade the performance of the target model from 76.20% to 41.83%. Our results demonstrate that TTA algorithms lacking a rigorous security assessment are unsuitable for deployment in real-life scenarios. As such, we advocate for the integration of defenses against test-time poisoning attacks into the design of TTA methods.