Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security最新文献

Network Anomaly Detection Using Transfer Learning Based on Auto-Encoders Loss Normalization 基于自编码器损失归一化的迁移学习网络异常检测

Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security Pub Date : 2021-11-15 DOI: 10.1145/3474369.3486869

Aviv Yehezkel, Eyal Elyashiv, Or Soffer

引用次数: 13

Investigating Labelless Drift Adaptation for Malware Detection 研究无标签漂移适应恶意软件检测

Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security Pub Date : 2021-11-15 DOI: 10.1145/3474369.3486873

Zeliang Kan, Feargus Pendlebury, Fabio Pierazzi, L. Cavallaro

{"title":"Investigating Labelless Drift Adaptation for Malware Detection","authors":"Zeliang Kan, Feargus Pendlebury, Fabio Pierazzi, L. Cavallaro","doi":"10.1145/3474369.3486873","DOIUrl":"https://doi.org/10.1145/3474369.3486873","url":null,"abstract":"The evolution of malware has long plagued machine learning-based detection systems, as malware authors develop innovative strategies to evade detection and chase profits. This induces concept drift as the test distribution diverges from the training, causing performance decay that requires constant monitoring and adaptation. In this work, we analyze the adaptation strategy used by DroidEvolver, a state-of-the-art learning system that self-updates using pseudo-labels to avoid the high overhead associated with obtaining a new ground truth. After removing sources of experimental bias present in the original evaluation, we identify a number of flaws in the generation and integration of these pseudo-labels, leading to a rapid onset of performance degradation as the model poisons itself. We propose DroidEvolver++, a more robust variant of DroidEvolver, to address these issues and highlight the role of pseudo-labels in addressing concept drift. We test the tolerance of the adaptation strategy versus different degrees of pseudo-label noise and propose the adoption of methods to ensure only high-quality pseudo-labels are used for updates. Ultimately, we conclude that the use of pseudo-labeling remains a promising solution to limitations on labeling capacity, but great care must be taken when designing update mechanisms to avoid negative feedback loops and self-poisoning which have catastrophic effects on performance.","PeriodicalId":411057,"journal":{"name":"Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114783237","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 9

SEAT 座位

Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security Pub Date : 2021-11-15 DOI: 10.1163/1574-9347_bnp_e1124640

Zhanyuan Zhang, Yizheng Chen, David A. Wagner

引用次数: 0

Automated Detection of Side Channels in Cryptographic Protocols: DROWN the ROBOTs! 加密协议中侧信道的自动检测:淹没机器人!

Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security Pub Date : 2021-11-15 DOI: 10.1145/3474369.3486868

J. P. Drees, Pritha Gupta, E. Hüllermeier, Tibor Jager, Alexander Konze, Claudia Priesterjahn, Arunselvan Ramaswamy, Juraj Somorovsky

{"title":"Automated Detection of Side Channels in Cryptographic Protocols: DROWN the ROBOTs!","authors":"J. P. Drees, Pritha Gupta, E. Hüllermeier, Tibor Jager, Alexander Konze, Claudia Priesterjahn, Arunselvan Ramaswamy, Juraj Somorovsky","doi":"10.1145/3474369.3486868","DOIUrl":"https://doi.org/10.1145/3474369.3486868","url":null,"abstract":"Currently most practical attacks on cryptographic protocols like TLS are based on side channels, such as padding oracles. Some well-known recent examples are DROWN, ROBOT and Raccoon (USENIX Security 2016, 2018, 2021). Such attacks are usually found by careful and time-consuming manual analysis by specialists. In this paper, we consider the question of how such attacks can be systematically detected and prevented before (large-scale) deployment. We propose a new, fully automated approach, which uses supervised learning to identify arbitrary patterns in network protocol traffic. In contrast to classical scanners, which search for known side channels, the detection of general patterns might detect new side channels, even unexpected ones, such as those from the ROBOT attack. To analyze this approach, we develop a tool to detect Bleichenbacher-like padding oracles in TLS server implementations, based on an ensemble of machine learning algorithms. We verify that the approach indeed detects known vulnerabilities successfully and reliably. The tool also provides detailed information about detected patterns to developers, to assist in removing a potential padding oracle. Due to the automation, the approach scales much better than manual analysis and could even be integrated with a CI/CD pipeline of a development environment, for example.","PeriodicalId":411057,"journal":{"name":"Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131396402","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security Pub Date : 2021-11-15 DOI: 10.1145/3474369.3486863

Zhanyuan Zhang, Yizheng Chen, David A. Wagner

引用次数: 13

Patch-based Defenses against Web Fingerprinting Attacks 基于补丁的Web指纹防御

Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security Pub Date : 2021-11-15 DOI: 10.1145/3474369.3486875

Shawn Shan, A. Bhagoji, Haitao Zheng, Ben Y. Zhao

{"title":"Patch-based Defenses against Web Fingerprinting Attacks","authors":"Shawn Shan, A. Bhagoji, Haitao Zheng, Ben Y. Zhao","doi":"10.1145/3474369.3486875","DOIUrl":"https://doi.org/10.1145/3474369.3486875","url":null,"abstract":"Anonymity systems like Tor are vulnerable to Website Fingerprinting (WF) attacks, where a local passive eavesdropper infers the victim's activity. WF attacks based on deep learning classifiers have successfully overcome numerous defenses. While recent defenses leveraging adversarial examples offer promise, these adversarial examples can only be computed after the network session has concluded, thus offering users little protection in practical settings. We propose Dolos, a system that modifies user network traffic in real time to successfully evade WF attacks. Dolos injects dummy packets into traffic traces by computing input-agnostic adversarial patches that disrupt the deep learning classifiers used in WF attacks. Patches are then applied to alter and protect user traffic in real time. Importantly, these patches are parameterized by a user-side secret, ensuring that attackers cannot use adversarial training to defeat Dolos. We experimentally demonstrate that Dolos provides >94% protection against state-of-the-art WF attacks under a variety of settings, including adaptive countermeasures. Dolos outperforms prior defenses both in terms of higher protection performance as well as lower bandwidth overhead. Finally, we show that Dolos is provably robust to any attack under specific, but realistic, assumptions on the setting in which the defense is deployed.","PeriodicalId":411057,"journal":{"name":"Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131863707","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 12

Session details: Session 2B: Machine Learning for Cybersecurity 会议详情:会议2B:网络安全的机器学习

Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security Pub Date : 2021-11-15 DOI: 10.1145/3494695

Ambra Demontis

引用次数: 0

Spying through Virtual Backgrounds of Video Calls 通过视频通话的虚拟背景进行监视

Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security Pub Date : 2021-11-15 DOI: 10.1145/3474369.3486870

Jan Malte Hilgefort, Dan Arp, Konrad Rieck

{"title":"Spying through Virtual Backgrounds of Video Calls","authors":"Jan Malte Hilgefort, Dan Arp, Konrad Rieck","doi":"10.1145/3474369.3486870","DOIUrl":"https://doi.org/10.1145/3474369.3486870","url":null,"abstract":"Video calls have become an essential part of today's business life, especially due to the Corona pandemic. Several industry branches enable their employees to work from home and collaborate via video conferencing services. While remote work offers benefits for health safety and personal mobility, it also poses privacy risks. Visual content is directly transmitted from the private living environment of employees to third parties, potentially exposing sensitive information. To counter this threat, video conferencing services support replacing the visible environment of a video call with a virtual background. This replacement, however, is imperfect, leaking tiny regions of the real background in video frames. In this paper, we explore how these leaks in virtual backgrounds can be exploited to reconstruct regions of the real environment. To this end, we build on recent techniques of computer vision and derive an approach capable of extracting and aggregating leaked pixels in a video call. In an empirical study with the services Zoom, Webex, and Google Meet, we can demonstrate that the exposed fragments of the reconstructed background are sufficient to spot different objects. From 114 video calls with virtual backgrounds, 35% enable to correctly identify objects in the environment. We conclude that virtual backgrounds provide only limited protection, and alternative defenses are needed.","PeriodicalId":411057,"journal":{"name":"Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126063796","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 2

Session details: Session 3: Privacy-Preserving Machine Learning 会议详情:会议3:隐私保护机器学习

Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security Pub Date : 2021-11-15 DOI: 10.1145/3494696

Yizheng Chen

引用次数: 0

Session details: Session 1: Adversarial Machine Learning 会议详情:会议1:对抗性机器学习

Proceedings of the 14th ACM Workshop on Artificial Intelligence and Security Pub Date : 2021-11-15 DOI: 10.1145/3494693

Nicholas Carlini

引用次数: 0