Cryptography and Security (cs.CR)

Abstract: Deep neural networks (DNNs) are well known to be vulnerable to adversarial examples (AEs). In addition, AEs have adversarial transferability, which means AEs generated for a source model can fool another black-box model (target model) with a non-trivial probability. In previous studies, it was confirmed that the vision transformer (ViT) is more robust against the property of adversarial transferability than convolutional neural network (CNN) models such as ConvMixer, and moreover encrypted ViT is more robust than ViT without any encryption. In this article, we propose a random ensemble of encrypted ViT models to achieve much more robust models. In experiments, the proposed scheme is verified to be more robust against not only black-box attacks but also white-box ones than convention methods.

Abstract: We propose cryptographic protocols to incorporate time provenance guarantees while meeting confidentiality and controlled sharing needs for research data. We demonstrate the efficacy of these mechanisms by developing and benchmarking a practical tool, GovernR, which furthermore takes into usability issues and is compatible with a popular open-sourced research data storage platform, Dataverse. In doing so, we identify and provide a solution addressing an important gap (though applicable to only niche use cases) in practical research data management.

Abstract: With the advance in malware technology, attackers create new ways to hide their malicious code from antivirus services. One way to obfuscate an attack is to use common files as cover to hide the malicious scripts, so the malware will look like a legitimate file. Although cutting-edge Artificial Intelligence and content signature exist, evasive malware successfully bypasses next-generation malware detection using advanced methods like steganography. Some of the files commonly used to hide malware are image files (e.g., JPEG). In addition, some malware use steganography to hide malicious scripts or sensitive data in images. Steganography in images is difficult to detect even with specialized tools. Image-based attacks try to attack the user's device using malicious payloads or utilize image steganography to hide sensitive data inside legitimate images and leak it outside the user's device. Therefore in this paper, we present a novel Image Content Disarm and Reconstruction (ICDR). Our ICDR system removes potential malware, with a zero trust approach, while maintaining high image quality and file usability. By extracting the image data, removing it from the rest of the file, and manipulating the image pixels, it is possible to disable or remove the hidden malware inside the file.

Abstract: Risk assessment plays a crucial role in ensuring the security and resilience of modern computer systems. Existing methods for conducting risk assessments often suffer from tedious and time-consuming processes, making it challenging to maintain a comprehensive overview of potential security issues. In this paper, we propose a novel approach that leverages attack graphs to enhance the efficiency and effectiveness of risk assessment. Attack graphs visually represent the various attack paths that adversaries can exploit within a system, enabling a systematic exploration of potential vulnerabilities. By extending attack graphs with capabilities to include countermeasures and consequences, they can be leveraged to constitute the complete risk assessment process. Our method offers a more streamlined and comprehensive analysis of system vulnerabilities, where system changes, or environment changes can easily be adapted and the issues exposing the highest risk can easily be identified. We demonstrate the effectiveness of our approach through a case study, as well as the applicability by combining existing risk assessment standards with our method. Our work aims to bridge the gap between risk assessment practices and evolving threat landscapes, offering an improved methodology for managing and mitigating risks in modern computer systems.

Abstract: We propose a new method to quantify the impact of cyber attacks in Cyber Physical Systems (CPSs). In particular, our method allows to identify the Design Parameter (DPs) affected due to a cyber attack launched on a different set of DPs in the same CPS. To achieve this, we adopt causal graphs to causally link DPs with each other and quantify the impact of one DP on another. Using SWaT, a real world testbed of a water treatment system, we demonstrate that causal graphs can be build in two ways: i) using domain knowledge of the control logic and the physical connectivity structure of the DPs, we call these causal domain graphs and ii) learning from operational data logs, we call these causal learnt graphs. We then compare these graphs when a same set of DPs is used. Our analysis shows a common set of edges between the causal domain graphs and the causal learnt graphs exists, which helps validate the causal learnt graphs. Additionally, we show that the learnt graphs can discover new causal relations, not initially considered in the domain graphs, that help significantly characterising the impact of the attack. We use causal domain graphs to estimate the parameters of the graphs, and the causal learnt graphs for causal inference. To learn the structure of the causal learnt graphs in all the six-stages of SWaT, we experiment with three learning algorithms: Peter Clarke (PC), Hill Climb (HC) search and Chow-Lie (CH). Finally, we demonstrate how causal graphs can be used to analyse the impact of cyber attacks by analysing nine well known cyber attacks on the SWaT test bed. We find that by using causal learnt graphs the DPs impacted by the attacks are correctly discovered with a probability greater than 0.9.

Abstract: This paper delves into the realm of ChatGPT, an AI-powered chatbot that utilizes topic modeling and reinforcement learning to generate natural responses. Although ChatGPT holds immense promise across various industries, such as customer service, education, mental health treatment, personal productivity, and content creation, it is essential to address its security, privacy, and ethical implications. By exploring the upgrade path from GPT-1 to GPT-4, discussing the model's features, limitations, and potential applications, this study aims to shed light on the potential risks of integrating ChatGPT into our daily lives. Focusing on security, privacy, and ethics issues, we highlight the challenges these concerns pose for widespread adoption. Finally, we analyze the open problems in these areas, calling for concerted efforts to ensure the development of secure and ethically sound large language models.