Xiaofeng Wang

Kun Wang, Guibin Zhang, Zhenhong Zhou et al. · mit

The remarkable success of Large Language Models (LLMs) has illuminated a promising pathway toward achieving Artificial General Intelligence for both academic and industrial communities, owing to their unprecedented performance across various applications. As LLMs continue to gain prominence in both research and commercial domains, their security and safety implications have become a growing concern, not only for researchers and corporations but also for every nation. Currently, existing surveys on LLM safety primarily focus on specific stages of the LLM lifecycle, e.g., deployment phase or fine-tuning phase, lacking a comprehensive understanding of the entire "lifechain" of LLMs. To address this gap, this paper introduces, for the first time, the concept of "full-stack" safety to systematically consider safety issues throughout the entire process of LLM training, deployment, and eventual commercialization. Compared to the off-the-shelf LLM safety surveys, our work demonstrates several distinctive advantages: (I) Comprehensive Perspective. We define the complete LLM lifecycle as encompassing data preparation, pre-training, post-training, deployment and final commercialization. To our knowledge, this represents the first safety survey to encompass the entire lifecycle of LLMs. (II) Extensive Literature Support. Our research is grounded in an exhaustive review of over 800+ papers, ensuring comprehensive coverage and systematic organization of security issues within a more holistic understanding. (III) Unique Insights. Through systematic literature analysis, we have developed reliable roadmaps and perspectives for each chapter. Our work identifies promising research directions, including safety in data generation, alignment techniques, model editing, and LLM-based agent systems. These insights provide valuable guidance for researchers pursuing future work in this field.

Zilong Lin, Jian Cui, Xiaojing Liao et al.

The underground exploitation of large language models (LLMs) for malicious services (i.e., Malla) is witnessing an uptick, amplifying the cyber threat landscape and posing questions about the trustworthiness of LLM technologies. However, there has been little effort to understand this new cybercrime, in terms of its magnitude, impact, and techniques. In this paper, we conduct the first systematic study on 212 real-world Mallas, uncovering their proliferation in underground marketplaces and exposing their operational modalities. Our study discloses the Malla ecosystem, revealing its significant growth and impact on today's public LLM services. Through examining 212 Mallas, we uncovered eight backend LLMs used by Mallas, along with 182 prompts that circumvent the protective measures of public LLM APIs. We further demystify the tactics employed by Mallas, including the abuse of uncensored LLMs and the exploitation of public LLM APIs through jailbreak prompts. Our findings enable a better understanding of the real-world exploitation of LLMs by cybercriminals, offering insights into strategies to counteract this cybercrime.

Lingzhi Yuan, Xinfeng Li, Chejian Xu et al.

Recent text-to-image (T2I) models have exhibited remarkable performance in generating high-quality images from text descriptions. However, these models are vulnerable to misuse, particularly generating not-safe-for-work (NSFW) content, such as sexually explicit, violent, political, and disturbing images, raising serious ethical concerns. In this work, we present PromptGuard, a novel content moderation technique that draws inspiration from the system prompt mechanism in large language models (LLMs) for safety alignment. Unlike LLMs, T2I models lack a direct interface for enforcing behavioral guidelines. Our key idea is to optimize a safety soft prompt that functions as an implicit system prompt within the T2I model's textual embedding space. This universal soft prompt (P*) directly moderates NSFW inputs, enabling safe yet realistic image generation without altering the inference efficiency or requiring proxy models. We further enhance its reliability and helpfulness through a divide-and-conquer strategy, which optimizes category-specific soft prompts and combines them into holistic safety guidance. Extensive experiments across five datasets demonstrate that PromptGuard effectively mitigates NSFW content generation while preserving high-quality benign outputs. PromptGuard achieves 3.8 times faster than prior content moderation methods, surpassing eight state-of-the-art defenses with an optimal unsafe ratio down to 5.84%.

Xiaofeng Wang, Yiming Wang, Wenhong Zhu et al.

Geometric ability is a significant challenge for large language models (LLMs) due to the need for advanced spatial comprehension and abstract thinking. Existing datasets primarily evaluate LLMs on their final answers, but they cannot truly measure their true understanding of geometric structures, as LLMs can arrive at correct answers by coincidence. To fill this gap, we introduce the GeomRel dataset, designed to evaluate LLMs' understanding of geometric structures by isolating the core step of geometric relationship identification in problem-solving. Using this benchmark, we conduct thorough evaluations of diverse LLMs and identify key limitations in understanding geometric structures. We further propose the Geometry Chain-of-Thought (GeoCoT) method, which enhances LLMs' ability to identify geometric relationships, resulting in significant performance improvements.

Weijie Liu, Hongbo Chen, XiaoFeng Wang et al.

As an emerging technique for confidential computing, trusted execution environment (TEE) receives a lot of attention. To better develop, deploy, and run secure applications on a TEE platform such as Intel's SGX, both academic and industrial teams have devoted much effort to developing reliable and convenient TEE containers. In this paper, we studied the isolation strategies of 15 existing TEE containers to protect secure applications from potentially malicious operating systems (OS) or untrusted applications, using a semi-automatic approach combining a feedback-guided analyzer with manual code review. Our analysis reveals the isolation protection each of these TEE containers enforces, and their security weaknesses. We observe that none of the existing TEE containers can fulfill the goal they set, due to various pitfalls in their design and implementation. We report the lessons learnt from our study for guiding the development of more secure containers, and further discuss the trend of TEE container designs. We also release our analyzer that helps evaluate the container middleware both from the enclave and from the kernel.

Weijie Liu, Wenhao Wang, Xiaofeng Wang et al.

A trusted execution environment (TEE) such as Intel Software Guard Extension (SGX) runs a remote attestation to prove to a data owner the integrity of the initial state of an enclave, including the program to operate on her data. For this purpose, the data-processing program is supposed to be open to the owner, so its functionality can be evaluated before trust can be established. However, increasingly there are application scenarios in which the program itself needs to be protected. So its compliance with privacy policies as expected by the data owner should be verified without exposing its code. To this end, this paper presents CAT, a new model for TEE-based confidential attestation. Our model is inspired by Proof-Carrying Code, where a code generator produces proof together with the code and a code consumer verifies the proof against the code on its compliance with security policies. Given that the conventional solutions do not work well under the resource-limited and TCB-frugal TEE, we propose a new design that allows an untrusted out-enclave generator to analyze the source code of a program when compiling it into binary and a trusted in-enclave consumer efficiently verifies the correctness of the instrumentation and the presence of other protection before running the binary. Our design strategically moves most of the workload to the code generator, which is responsible for producing well-formatted and easy-to-check code, while keeping the consumer simple. Also, the whole consumer can be made public and verified through a conventional attestation. We implemented this model on Intel SGX and demonstrate that it introduces a very small part of TCB. We also thoroughly evaluated its performance on micro- and macro- benchmarks and real-world applications, showing that the new design only incurs a small overhead when enforcing several categories of security policies.

Wenhao Wang, Yichen Jiang, Qintao Shen et al.

It has been a long standing problem to securely outsource computation tasks to an untrusted party with integrity and confidentiality guarantees. While fully homomorphic encryption (FHE) is a promising technique that allows computations performed on the encrypted data, it suffers from a significant slow down to the computation. In this paper we propose a hybrid solution that uses the latest hardware Trusted Execution Environments (TEEs) to assist FHE by moving the bootstrapping step, which is one of the major obstacles in designing practical FHE schemes, to a secured SGX enclave. TEEFHE, the hybrid system we designed, makes it possible for homomorphic computations to be performed on smaller ciphertext and secret key, providing better performance and lower memory consumption. We make an effort to mitigate side channel leakages within SGX by making the memory access patterns totally independent from the secret information. The evaluation shows that TEEFHE effectively improves the software only FHE schemes in terms of both time and space.

Jianping Zhu, Rui Hou, XiaoFeng Wang et al.

There is an urgent demand for privacy-preserving techniques capable of supporting compute and data intensive (CDI) computing in the era of big data. However, none of existing TEEs can truly support CDI computing tasks, as CDI requires high throughput accelerators like GPU and TPU but TEEs do not offer security protection of such accelerators. This paper present HETEE (Heterogeneous TEE), the first design of TEE capable of strongly protecting heterogeneous computing with unsecure accelerators. HETEE is uniquely constructed to work with today's servers, and does not require any changes for existing commercial CPUs or accelerators. The key idea of our design runs security controller as a stand-alone computing system to dynamically adjust the boundary of between secure and insecure worlds through the PCIe switches, rendering the control of an accelerator to the host OS when it is not needed for secure computing, and shifting it back when it is. The controller is the only trust unit in the system and it runs the custom OS and accelerator runtimes, together with the encryption, authentication and remote attestation components. The host server and other computing systems communicate with controller through an in memory task queue that accommodates the computing tasks offloaded to HETEE, in the form of encrypted and signed code and data. Also, HETEE offers a generic and efficient programming model to the host CPU. We have implemented the HETEE design on a hardware prototype system, and evaluated it with large-scale Neural Networks inference and training tasks. Our evaluations show that HETEE can easily support such secure computing tasks and only incurs a 12.34% throughput overhead for inference and 9.87% overhead for training on average.

Arthur Evensen, Kai-Kristian Kemell, Xiaofeng Wang et al.

Software Engineering practitioners work using highly diverse methods and practices, and general theories in software engineering are lacking. One attempt at creating a common ground in the area of software engineering methodologies has been the Essence Theory of Software Engineering, which can be considered a method-agnostic project management tool for software engineering. Essence supports the use of any development practices and provides a framework for building a suitable method for any software engineering context. However, Essence presently suffers from low practitioner adoption that is partially considered to be caused by a lack of proper tooling. In this paper, we present Essencery, a tool for essentializing software engineering methods and practices using the Essence graphical syntax. Essencery aims to facilitate adoption of Essence among potential future users. We present an empirical evaluation of the tool by means of a qualitative, quasi-formal experiment and, based on the experiment, confirm that the tool is easy to use and useful for its intended purpose.

Xuejing Yuan, Yuxuan Chen, Yue Zhao et al.

The popularity of ASR (automatic speech recognition) systems, like Google Voice, Cortana, brings in security concerns, as demonstrated by recent attacks. The impacts of such threats, however, are less clear, since they are either less stealthy (producing noise-like voice commands) or requiring the physical presence of an attack device (using ultrasound). In this paper, we demonstrate that not only are more practical and surreptitious attacks feasible but they can even be automatically constructed. Specifically, we find that the voice commands can be stealthily embedded into songs, which, when played, can effectively control the target system through ASR without being noticed. For this purpose, we developed novel techniques that address a key technical challenge: integrating the commands into a song in a way that can be effectively recognized by ASR through the air, in the presence of background noise, while not being detected by a human listener. Our research shows that this can be done automatically against real world ASR applications. We also demonstrate that such CommanderSongs can be spread through Internet (e.g., YouTube) and radio, potentially affecting millions of ASR users. We further present a new mitigation technique that controls this threat.

Shuaike Dong, Menghao Li, Wenrui Diao et al.

In this paper, we seek to better understand Android obfuscation and depict a holistic view of the usage of obfuscation through a large-scale investigation in the wild. In particular, we focus on four popular obfuscation approaches: identifier renaming, string encryption, Java reflection, and packing. To obtain the meaningful statistical results, we designed efficient and lightweight detection models for each obfuscation technique and applied them to our massive APK datasets (collected from Google Play, multiple third-party markets, and malware databases). We have learned several interesting facts from the result. For example, malware authors use string encryption more frequently, and more apps on third-party markets than Google Play are packed. We are also interested in the explanation of each finding. Therefore we carry out in-depth code analysis on some Android apps after sampling. We believe our study will help developers select the most suitable obfuscation approach, and in the meantime help researchers improve code analysis systems in the right direction.

Wenhao Wang, Guoxing Chen, Xiaorui Pan et al.

Side-channel risks of Intel's SGX have recently attracted great attention. Under the spotlight is the newly discovered page-fault attack, in which an OS-level adversary induces page faults to observe the page-level access patterns of a protected process running in an SGX enclave. With almost all proposed defense focusing on this attack, little is known about whether such efforts indeed raise the bar for the adversary, whether a simple variation of the attack renders all protection ineffective, not to mention an in-depth understanding of other attack surfaces in the SGX system. In the paper, we report the first step toward systematic analyses of side-channel threats that SGX faces, focusing on the risks associated with its memory management. Our research identifies 8 potential attack vectors, ranging from TLB to DRAM modules. More importantly, we highlight the common misunderstandings about SGX memory side channels, demonstrating that high frequent AEXs can be avoided when recovering EdDSA secret key through a new page channel and fine-grained monitoring of enclave programs (at the level of 64B) can be done through combining both cache and cross-enclave DRAM channels. Our findings reveal the gap between the ongoing security research on SGX and its side-channel weaknesses, redefine the side-channel threat model for secure enclaves, and can provoke a discussion on when to use such a system and how to use it securely.

Nan Zhang, Soteris Demetriou, Xianghang Mi et al.

Inspired by the boom of the consumer IoT market, many device manufacturers, start-up companies and technology giants have jumped into the space. Unfortunately, the exciting utility and rapid marketization of IoT, come at the expense of privacy and security. Industry reports and academic work have revealed many attacks on IoT systems, resulting in privacy leakage, property loss and large-scale availability problems. To mitigate such threats, a few solutions have been proposed. However, it is still less clear what are the impacts they can have on the IoT ecosystem. In this work, we aim to perform a comprehensive study on reported attacks and defenses in the realm of IoT aiming to find out what we know, where the current studies fall short and how to move forward. To this end, we first build a toolkit that searches through massive amount of online data using semantic analysis to identify over 3000 IoT-related articles. Further, by clustering such collected data using machine learning technologies, we are able to compare academic views with the findings from industry and other sources, in an attempt to understand the gaps between them, the trend of the IoT security risks and new problems that need further attention. We systemize this process, by proposing a taxonomy for the IoT ecosystem and organizing IoT security into five problem areas. We use this taxonomy as a beacon to assess each IoT work across a number of properties we define. Our assessment reveals that relevant security and privacy problems are far from solved. We discuss how each proposed solution can be applied to a problem area and highlight their strengths, assumptions and constraints. We stress the need for a security framework for IoT vendors and discuss the trend of shifting security liability to external or centralized entities. We also identify open research problems and provide suggestions towards a secure IoT ecosystem.

Soteris Demetriou, Nan Zhang, Yeonjoon Lee et al.

A new development of smart-home systems is to use mobile apps to control IoT devices across a Home Area Network (HAN). Those systems tend to rely on the Wi-Fi router to authenticate other devices; as verified in our study, IoT vendors tend to trust all devices connected to the HAN. This treatment exposes them to the attack from malicious apps, particularly those running on authorized phones, which the router does not have information to control, as confirmed in our measurement study. Mitigating this threat cannot solely rely on IoT manufacturers, which may need to change the hardware on the devices to support encryption, increasing the cost of the device, or software developers who we need to trust to implement security correctly. In this work, we present a new technique to control the communication between the IoT devices and their apps in a unified, backward-compatible way. Our approach, called Hanguard, does not require any changes to the IoT devices themselves, the IoT apps or the OS of the participating phones. Hanguard achieves a fine-grained, per-app protection through bridging the OS-level situation awareness and the router-level per-flow control: each phone runs a non-system userspace Monitor app to identify the party that attempts to access the protected IoT device and inform the router through a control plane of its access decision; the router enforces the decision on the data plane after verifying whether the phone should be allowed to talk to the device. Hanguard uses a role-based access control (RBAC) schema which leverages type enforcement (TE) and multi-category security (MCS) primitives to define highly flexible access control rules. We implemented our design over both Android and iOS (>95% of mobile OS market share) and a popular router. Our study shows that Hanguard is both efficient and effective in practice.

Luyi Xing, Xiaolong Bai, Tongxin Li et al.

On modern operating systems, applications under the same user are separated from each other, for the purpose of protecting them against malware and compromised programs. Given the complexity of today's OSes, less clear is whether such isolation is effective against different kind of cross-app resource access attacks (called XARA in our research). To better understand the problem, on the less-studied Apple platforms, we conducted a systematic security analysis on MAC OS~X and iOS. Our research leads to the discovery of a series of high-impact security weaknesses, which enable a sandboxed malicious app, approved by the Apple Stores, to gain unauthorized access to other apps' sensitive data. More specifically, we found that the inter-app interaction services, including the keychain, WebSocket and NSConnection on OS~X and URL Scheme on the MAC OS and iOS, can all be exploited by the malware to steal such confidential information as the passwords for iCloud, email and bank, and the secret token of Evernote. Further, the design of the app sandbox on OS~X was found to be vulnerable, exposing an app's private directory to the sandboxed malware that hijacks its Apple Bundle ID. As a result, sensitive user data, like the notes and user contacts under Evernote and photos under WeChat, have all been disclosed. Fundamentally, these problems are caused by the lack of app-to-app and app-to-OS authentications. To better understand their impacts, we developed a scanner that automatically analyzes the binaries of MAC OS and iOS apps to determine whether proper protection is missing in their code. Running it on hundreds of binaries, we confirmed the pervasiveness of the weaknesses among high-impact Apple apps. Since the issues may not be easily fixed, we built a simple program that detects exploit attempts on OS~X, helping protect vulnerable apps before the problems can be fully addressed.