Ahmad-Reza Sadeghi

Maximilian Thang, Lichao Wu, Sasha Behrouzi et al.

Large language models (LLMs) are increasingly used for code generation in fast, informal development workflows, often referred to as vibe coding, where speed and convenience are prioritized, and security requirements are rarely made explicit. In this setting, models frequently produce functionally correct but insecure code, creating a growing security risk. Existing approaches to improving code security rely on full-parameter fine-tuning or parameter-efficient adaptations, which are either costly and prone to catastrophic forgetting or operate at coarse granularity with limited interpretability and control. We present GoodVibe, a neuron-level framework for improving the security of code language models by default. GoodVibe is based on the key insight that security-relevant reasoning is localized to a small subset of neurons. We identify these neurons using gradient-based attribution from a supervised security task and perform neuron-selective fine-tuning that updates only this security-critical subspace. To further reduce training cost, we introduce activation-driven neuron clustering, enabling structured updates with minimal overhead. We evaluate GoodVibe on six LLMs across security-critical programming languages, including C++, Java, Swift, and Go. GoodVibe substantially improves the security of generated code while preserving general model utility, achieving up to a 2.5x improvement over base models, achieving performance competitive with full fine-tuning while using over 4,700x fewer trainable parameters, and reducing training computation by more than 3.6x compared to the parameter-efficient baseline (LoRA). Our results demonstrate that neuron-level optimization offers an effective and scalable approach to securing code generation without sacrificing generality.

Alessandro Pegoraro, Kavita Kumari, Hossein Fereidooni et al.

ChatGPT has become a global sensation. As ChatGPT and other Large Language Models (LLMs) emerge, concerns of misusing them in various ways increase, such as disseminating fake news, plagiarism, manipulating public opinion, cheating, and fraud. Hence, distinguishing AI-generated from human-generated becomes increasingly essential. Researchers have proposed various detection methodologies, ranging from basic binary classifiers to more complex deep-learning models. Some detection techniques rely on statistical characteristics or syntactic patterns, while others incorporate semantic or contextual information to improve accuracy. The primary objective of this study is to provide a comprehensive and contemporary assessment of the most recent techniques in ChatGPT detection. Additionally, we evaluated other AI-generated text detection tools that do not specifically claim to detect ChatGPT-generated content to assess their performance in detecting ChatGPT-generated content. For our evaluation, we have curated a benchmark dataset consisting of prompts from ChatGPT and humans, including diverse questions from medical, open Q&A, and finance domains and user-generated responses from popular social networking platforms. The dataset serves as a reference to assess the performance of various techniques in detecting ChatGPT-generated content. Our evaluation results demonstrate that none of the existing methods can effectively detect ChatGPT-generated content.

Kavita Kumari, Phillip Rieger, Hossein Fereidooni et al.

Federated learning (FL) allows participants to jointly train a machine learning model without sharing their private data with others. However, FL is vulnerable to poisoning attacks such as backdoor attacks. Consequently, a variety of defenses have recently been proposed, which have primarily utilized intermediary states of the global model (i.e., logits) or distance of the local models (i.e., L2-norm) from the global model to detect malicious backdoors. However, as these approaches directly operate on client updates, their effectiveness depends on factors such as clients' data distribution or the adversary's attack strategies. In this paper, we introduce a novel and more generic backdoor defense framework, called BayBFed, which proposes to utilize probability distributions over client updates to detect malicious updates in FL: it computes a probabilistic measure over the clients' updates to keep track of any adjustments made in the updates, and uses a novel detection algorithm that can leverage this probabilistic measure to efficiently detect and filter out malicious updates. Thus, it overcomes the shortcomings of previous approaches that arise due to the direct usage of client updates; as our probabilistic measure will include all aspects of the local client training strategies. BayBFed utilizes two Bayesian Non-Parametric extensions: (i) a Hierarchical Beta-Bernoulli process to draw a probabilistic measure given the clients' updates, and (ii) an adaptation of the Chinese Restaurant Process (CRP), referred by us as CRP-Jensen, which leverages this probabilistic measure to detect and filter out malicious updates. We extensively evaluate our defense approach on five benchmark datasets: CIFAR10, Reddit, IoT intrusion detection, MNIST, and FMNIST, and show that it can effectively detect and eliminate malicious updates in FL without deteriorating the benign performance of the global model.

Phillip Rieger, Marco Chilese, Reham Mohamed et al.

IoT application domains, device diversity and connectivity are rapidly growing. IoT devices control various functions in smart homes and buildings, smart cities, and smart factories, making these devices an attractive target for attackers. On the other hand, the large variability of different application scenarios and inherent heterogeneity of devices make it very challenging to reliably detect abnormal IoT device behaviors and distinguish these from benign behaviors. Existing approaches for detecting attacks are mostly limited to attacks directly compromising individual IoT devices, or, require predefined detection policies. They cannot detect attacks that utilize the control plane of the IoT system to trigger actions in an unintended/malicious context, e.g., opening a smart lock while the smart home residents are absent. In this paper, we tackle this problem and propose ARGUS, the first self-learning intrusion detection system for detecting contextual attacks on IoT environments, in which the attacker maliciously invokes IoT device actions to reach its goals. ARGUS monitors the contextual setting based on the state and actions of IoT devices in the environment. An unsupervised Deep Neural Network (DNN) is used for modeling the typical contextual device behavior and detecting actions taking place in abnormal contextual settings. This unsupervised approach ensures that ARGUS is not restricted to detecting previously known attacks but is also able to detect new attacks. We evaluated ARGUS on heterogeneous real-world smart-home settings and achieve at least an F1-Score of 99.64% for each setup, with a false positive rate (FPR) of at most 0.03%.

Pier Paolo Tricomi, Marco Chilese, Mauro Conti et al.

With 1.3 billion users, Instagram (IG) has also become a business tool. IG influencer marketing, expected to generate $33.25 billion in 2022, encourages companies and influencers to create trending content. Various methods have been proposed for predicting a post's popularity, i.e., how much engagement (e.g., Likes) it will generate. However, these methods are limited: first, they focus on forecasting the likes, ignoring the number of comments, which became crucial in 2021. Secondly, studies often use biased or limited data. Third, researchers focused on Deep Learning models to increase predictive performance, which are difficult to interpret. As a result, end-users can only estimate engagement after a post is created, which is inefficient and expensive. A better approach is to generate a post based on what people and IG like, e.g., by following guidelines. In this work, we uncover part of the underlying mechanisms driving IG engagement. To achieve this goal, we rely on statistical analysis and interpretable models rather than Deep Learning (black-box) approaches. We conduct extensive experiments using a worldwide dataset of 10 million posts created by 34K global influencers in nine different categories. With our simple yet powerful algorithms, we can predict engagement up to 94% of F1-Score, making us comparable and even superior to Deep Learning-based method. Furthermore, we propose a novel unsupervised algorithm for finding highly engaging topics on IG. Thanks to our interpretable approaches, we conclude by outlining guidelines for creating successful posts.

Lichao Wu, Sasha Behrouzi, Mohamadreza Rostami et al.

Safety alignment is critical for the ethical deployment of large language models (LLMs), guiding them to avoid generating harmful or unethical content. Current alignment techniques, such as supervised fine-tuning and reinforcement learning from human feedback, remain fragile and can be bypassed by carefully crafted adversarial prompts. Unfortunately, such attacks rely on trial and error, lack generalizability across models, and are constrained by scalability and reliability. This paper presents NeuroStrike, a novel and generalizable attack framework that exploits a fundamental vulnerability introduced by alignment techniques: the reliance on sparse, specialized safety neurons responsible for detecting and suppressing harmful inputs. We apply NeuroStrike to both white-box and black-box settings: In the white-box setting, NeuroStrike identifies safety neurons through feedforward activation analysis and prunes them during inference to disable safety mechanisms. In the black-box setting, we propose the first LLM profiling attack, which leverages safety neuron transferability by training adversarial prompt generators on open-weight surrogate models and then deploying them against black-box and proprietary targets. We evaluate NeuroStrike on over 20 open-weight LLMs from major LLM developers. By removing less than 0.6% of neurons in targeted layers, NeuroStrike achieves an average attack success rate (ASR) of 76.9% using only vanilla malicious prompts. Moreover, Neurostrike generalizes to four multimodal LLMs with 100% ASR on unsafe image inputs. Safety neurons transfer effectively across architectures, raising ASR to 78.5% on 11 fine-tuned models and 77.7% on five distilled models. The black-box LLM profiling attack achieves an average ASR of 63.7% across five black-box models, including the Google Gemini family.

Phillip Rieger, Torsten Krauß, Markus Miettinen et al.

Federated Learning (FL) is a promising approach enabling multiple clients to train Deep Neural Networks (DNNs) collaboratively without sharing their local training data. However, FL is susceptible to backdoor (or targeted poisoning) attacks. These attacks are initiated by malicious clients who seek to compromise the learning process by introducing specific behaviors into the learned model that can be triggered by carefully crafted inputs. Existing FL safeguards have various limitations: They are restricted to specific data distributions or reduce the global model accuracy due to excluding benign models or adding noise, are vulnerable to adaptive defense-aware adversaries, or require the server to access local models, allowing data inference attacks. This paper presents a novel defense mechanism, CrowdGuard, that effectively mitigates backdoor attacks in FL and overcomes the deficiencies of existing techniques. It leverages clients' feedback on individual models, analyzes the behavior of neurons in hidden layers, and eliminates poisoned models through an iterative pruning scheme. CrowdGuard employs a server-located stacked clustering scheme to enhance its resilience to rogue client feedback. The evaluation results demonstrate that CrowdGuard achieves a 100% True-Positive-Rate and True-Negative-Rate across various scenarios, including IID and non-IID data distributions. Additionally, CrowdGuard withstands adaptive adversaries while preserving the original performance of protected models. To ensure confidentiality, CrowdGuard uses a secure and privacy-preserving architecture leveraging Trusted Execution Environments (TEEs) on both client and server sides.

Martin Herrmann, Oussama Draissi, Christian Niesler et al.

Microarchitectural vulnerabilities increasingly undermine the assumption that hardware can be treated as a reliable root of trust. Prevention mechanisms often lag behind evolving attack techniques, leaving deployed systems unable to assume continued trustworthiness. We propose a shift from prevention to detection through microarchitectural-aware remote attestation. As a first instantiation of this idea, we present HammerWatch, a Rowhammer-aware remote attestation protocol that enables an external verifier to assess whether a system exhibits hardware-induced disturbance behavior. HammerWatch leverages memory-level evidence available on commodity platforms, specifically Machine-Check Exceptions (MCEs) from ECC DRAM and counter-based indicators from Per-Row Activation Counting (PRAC), and protects these measurements against kernel-level adversaries using TPM-anchored hash chains. We implement HammerWatch on commodity hardware and evaluate it on 20000 simulated benign and malicious access patterns. Our results show that the verifier reliably distinguishes Rowhammer-like behavior from benign operation under conservative heuristics, demonstrating that detection-oriented attestation is feasible and can complement incomplete prevention mechanisms

Kavita Kumari, Alessandro Pegoraro, Hossein Fereidooni et al.

The potential misuse of ChatGPT and other Large Language Models (LLMs) has raised concerns regarding the dissemination of false information, plagiarism, academic dishonesty, and fraudulent activities. Consequently, distinguishing between AI-generated and human-generated content has emerged as an intriguing research topic. However, current text detection methods lack precision and are often restricted to specific tasks or domains, making them inadequate for identifying content generated by ChatGPT. In this paper, we propose an effective ChatGPT detector named DEMASQ, which accurately identifies ChatGPT-generated content. Our method addresses two critical factors: (i) the distinct biases in text composition observed in human- and machine-generated content and (ii) the alterations made by humans to evade previous detection methods. DEMASQ is an energy-based detection model that incorporates novel aspects, such as (i) optimization inspired by the Doppler effect to capture the interdependence between input text embeddings and output labels, and (ii) the use of explainable AI techniques to generate diverse perturbations. To evaluate our detector, we create a benchmark dataset comprising a mixture of prompts from both ChatGPT and humans, encompassing domains such as medical, open Q&A, finance, wiki, and Reddit. Our evaluation demonstrates that DEMASQ achieves high accuracy in identifying content generated by ChatGPT.

Oussama Draissi, Mark Günzel, Ahmad-Reza Sadeghi et al.

WebAssembly's (Wasm) monolithic linear memory model facilitates memory corruption attacks that can escalate to cross-site scripting in browsers or go undetected when a malicious host tampers with a module's state. Existing defenses rely on invasive binary instrumentation or custom runtimes, and do not address runtime integrity verification under an adversarial host model. We present Walma, a framework for WebAssembly Linear Memory Attestation that leverages machine learning to detect memory corruption and external tampering by classifying memory snapshots. We evaluate Walma on six real-world CVE-affected applications across three verification backends (cpu-wasm, cpu-tch, gpu) and three instrumentation policies. Our results demonstrate that CNN-based classification can effectively detect memory corruption in applications with structured memory layouts, with coarse-grained boundary checks incurring as low as 1.07x overhead, while fine-grained monitoring introduces higher (1.5x--1.8x) but predictable costs. Our evaluation quantifies the accuracy and overhead trade-offs across deployment configurations, demonstrating the practical feasibility of ML-based memory attestation for WebAssembly.

Sahar Abdelnabi, Chris Hicks, Konrad Rieck et al.

The benchmarks used to evaluate AI agents in security-critical roles suffer from crucial weaknesses. Building on recent empirical evidence, we characterize three core challenges that undermine security evaluations: benchmark vulnerabilities, temporal staleness, and runtime uncertainty. We then outline practical directions toward building more robust and trustworthy evaluation frameworks.

Mohamadreza Rostami, Marco Chilese, Shaza Zeitouni et al.

Modern computing systems heavily rely on hardware as the root of trust. However, their increasing complexity has given rise to security-critical vulnerabilities that cross-layer at-tacks can exploit. Traditional hardware vulnerability detection methods, such as random regression and formal verification, have limitations. Random regression, while scalable, is slow in exploring hardware, and formal verification techniques are often concerned with manual effort and state explosions. Hardware fuzzing has emerged as an effective approach to exploring and detecting security vulnerabilities in large-scale designs like modern processors. They outperform traditional methods regarding coverage, scalability, and efficiency. However, state-of-the-art fuzzers struggle to achieve comprehensive coverage of intricate hardware designs within a practical timeframe, often falling short of a 70% coverage threshold. We propose a novel ML-based hardware fuzzer, ChatFuzz, to address this challenge. Ourapproach leverages LLMs like ChatGPT to understand processor language, focusing on machine codes and generating assembly code sequences. RL is integrated to guide the input generation process by rewarding the inputs using code coverage metrics. We use the open-source RISCV-based RocketCore processor as our testbed. ChatFuzz achieves condition coverage rate of 75% in just 52 minutes compared to a state-of-the-art fuzzer, which requires a lengthy 30-hour window to reach a similar condition coverage. Furthermore, our fuzzer can attain 80% coverage when provided with a limited pool of 10 simulation instances/licenses within a 130-hour window. During this time, it conducted a total of 199K test cases, of which 6K produced discrepancies with the processor's golden model. Our analysis identified more than 10 unique mismatches, including two new bugs in the RocketCore and discrepancies from the RISC-V ISA Simulator.

Sasha Behrouzi, Lichao Wu, Mohamadreza Rostami et al.

Safety alignment is essential for the responsible deployment of large language models (LLMs). Yet, existing approaches often rely on heavyweight fine-tuning that is costly to update, audit, and maintain across model families. Full fine-tuning incurs substantial computational and storage overhead, while parameter-efficient methods such as LoRA trade efficiency for inconsistent safety gains and sensitivity to design choices. Safety intervention mechanisms such as circuit breakers reduce unsafe outputs without modifying model weights, but do not directly shape or preserve the internal representations that govern safety behavior. These limitations hinder rapid and reliable safety updates, particularly in settings where models evolve frequently or must adapt to new policies and domains. We present NeST, a lightweight, structure-aware safety alignment framework that strengthens refusal behavior by selectively adapting a small subset of safety-relevant neurons while freezing the remainder of the model. NeST aligns parameter updates with the internal organization of safety behavior by clustering functionally coherent safety neurons and enforcing shared updates within each cluster, enabling targeted and stable safety adaptation without broad model modification or inference-time overhead. We benchmark NeST against three dominant baselines: full fine-tuning, LoRA-based fine-tuning, and circuit breakers across 10 open-weight LLMs spanning multiple model families and sizes. Across all evaluated models, NeST reduces the attack success rate from an average of 44.5% to 4.36%, corresponding to a 90.2% reduction in unsafe generations, while requiring only 0.44 million trainable parameters on average. This amounts to a 17,310x decrease in updated parameters compared to full fine-tuning and a 9.25x reduction relative to LoRA, while consistently achieving stronger safety performance for alignment.

Mert Cihangiroglu, Alessandro Pegoraro, Phillip Rieger et al.

Split learning (SL) enables collaborative training by partitioning a neural network across clients and a central server, but the cut-layer interface introduces a key challenge: high-dimensional activations incur substantial communication overhead while exposing representations vulnerable to reconstruction attacks. Existing approaches typically address efficiency or privacy in isolation, relying on additional mechanisms such as sparsification, quantization, or noise injection. We propose LightSplit, which limits information exposure and reduces communication overhead by applying a lightweight fixed orthogonal random projection at the cut layer. Based on Shannon's information theory, this projection acts as an information bottleneck that restricts instance-specific information and suppresses exploitable per-sample signals. By transmitting low-dimensional projections instead of raw activations, the server operates on lifted representations without requiring architectural modifications, ensuring compatibility with existing SL architectures. By avoiding additional trainable components on the client, the method remains lightweight and suitable for edge devices while preserving end-to-end differentiability via exact gradient propagation. As the projection is non-invertible, part of the original representation is irreversibly discarded at the client, LightSplit reduces the information available for reconstruction and limits information exposure. We extensively evaluate LightSplit on state-of-the-art benchmarks in both IID and non-IID settings across varying projection dimensions and client scales. Our results show that the method retains more than 95% of the baseline accuracy at up to 32x reduction in transmitted dimensionality while maintaining stable training dynamics.

Aakash Tyagi, Addison Crump, Ahmad-Reza Sadeghi et al.

The increasing complexity of modern processors poses many challenges to existing hardware verification tools and methodologies for detecting security-critical bugs. Recent attacks on processors have shown the fatal consequences of uncovering and exploiting hardware vulnerabilities. Fuzzing has emerged as a promising technique for detecting software vulnerabilities. Recently, a few hardware fuzzing techniques have been proposed. However, they suffer from several limitations, including non-applicability to commonly used Hardware Description Languages (HDLs) like Verilog and VHDL, the need for significant human intervention, and inability to capture many intrinsic hardware behaviors, such as signal transitions and floating wires. In this paper, we present the design and implementation of a novel hardware fuzzer, TheHuzz, that overcomes the aforementioned limitations and significantly improves the state of the art. We analyze the intrinsic behaviors of hardware designs in HDLs and then measure the coverage metrics that model such behaviors. TheHuzz generates assembly-level instructions to increase the desired coverage values, thereby finding many hardware bugs that are exploitable from software. We evaluate TheHuzz on four popular open-source processors and achieve 1.98x and 3.33x the speed compared to the industry-standard random regression approach and the state-of-the-art hardware fuzzer, DiffuzRTL, respectively. Using TheHuzz, we detected 11 bugs in these processors, including 8 new vulnerabilities, and we demonstrate exploits using the detected bugs. We also show that TheHuzz overcomes the limitations of formal verification tools from the semiconductor industry by comparing its findings to those discovered by the Cadence JasperGold tool.

Andrew Paverd, Sandeep Tamrakar, Hoang Long Nguyen et al.

Cloud storage services like Dropbox and Google Drive are widely used by individuals and businesses. Two attractive features of these services are 1) the automatic synchronization of files between multiple client devices and 2) the possibility to share files with other users. However, privacy of cloud data is a growing concern for both individuals and businesses. Encrypting data on the client-side before uploading it is an effective privacy safeguard, but it requires all client devices to have the decryption key. Current solutions derive these keys solely from user-chosen passwords, which have low entropy and are easily guessed. We present OmniShare, the first scheme to allow client-side encryption with high-entropy keys whilst providing an intuitive key distribution mechanism to enable access from multiple client devices. Instead of passwords, we use low bandwidth uni-directional out-of-band (OOB) channels, such as QR codes, to authenticate new devices. To complement these OOB channels, the cloud storage itself is used as a communication channel between devices in our protocols. We rely on a directory-based key hierarchy with individual file keys to limit the consequences of key compromise and allow efficient sharing of files without requiring re-encryption. OmniShare is open source software and currently available for Android and Windows with other platforms in development. We describe the design and implementation of OmniShare, and explain how we evaluated its security using formal methods, its performance via real-world benchmarks, and its usability through a cognitive walkthrough.

Hossein Fereidooni, Alessandro Pegoraro, Phillip Rieger et al.

Federated learning (FL) is a collaborative learning paradigm allowing multiple clients to jointly train a model without sharing their training data. However, FL is susceptible to poisoning attacks, in which the adversary injects manipulated model updates into the federated model aggregation process to corrupt or destroy predictions (untargeted poisoning) or implant hidden functionalities (targeted poisoning or backdoors). Existing defenses against poisoning attacks in FL have several limitations, such as relying on specific assumptions about attack types and strategies or data distributions or not sufficiently robust against advanced injection techniques and strategies and simultaneously maintaining the utility of the aggregated model. To address the deficiencies of existing defenses, we take a generic and completely different approach to detect poisoning (targeted and untargeted) attacks. We present FreqFed, a novel aggregation mechanism that transforms the model updates (i.e., weights) into the frequency domain, where we can identify the core frequency components that inherit sufficient information about weights. This allows us to effectively filter out malicious updates during local training on the clients, regardless of attack types, strategies, and clients' data distributions. We extensively evaluate the efficiency and effectiveness of FreqFed in different application domains, including image classification, word prediction, IoT intrusion detection, and speech recognition. We demonstrate that FreqFed can mitigate poisoning attacks effectively with a negligible impact on the utility of the aggregated model.

Marco Chilese, Richard Mitev, Meni Orenbach et al.

Control-Flow Attestation (CFA) is a security service that allows an entity (verifier) to verify the integrity of code execution on a remote computer system (prover). Existing CFA schemes suffer from impractical assumptions, such as requiring access to the prover's internal state (e.g., memory or code), the complete Control-Flow Graph (CFG) of the prover's software, large sets of measurements, or tailor-made hardware. Moreover, current CFA schemes are inadequate for attesting embedded systems due to their high computational overhead and resource usage. In this paper, we overcome the limitations of existing CFA schemes for embedded devices by introducing RAGE, a novel, lightweight CFA approach with minimal requirements. RAGE can detect Code Reuse Attacks (CRA), including control- and non-control-data attacks. It efficiently extracts features from one execution trace and leverages Unsupervised Graph Neural Networks (GNNs) to identify deviations from benign executions. The core intuition behind RAGE is to exploit the correspondence between execution trace, execution graph, and execution embeddings to eliminate the unrealistic requirement of having access to a complete CFG. We evaluate RAGE on embedded benchmarks and demonstrate that (i) it detects 40 real-world attacks on embedded software; (ii) Further, we stress our scheme with synthetic return-oriented programming (ROP) and data-oriented programming (DOP) attacks on the real-world embedded software benchmark Embench, achieving 98.03% (ROP) and 91.01% (DOP) F1-Score while maintaining a low False Positive Rate of 3.19%; (iii) Additionally, we evaluate RAGE on OpenSSL, used by millions of devices and achieve 97.49% and 84.42% F1-Score for ROP and DOP attack detection, with an FPR of 5.47%.

Phillip Rieger, Alessandro Pegoraro, Kavita Kumari et al.

Split Learning (SL) is a distributed deep learning approach enabling multiple clients and a server to collaboratively train and infer on a shared deep neural network (DNN) without requiring clients to share their private local data. The DNN is partitioned in SL, with most layers residing on the server and a few initial layers and inputs on the client side. This configuration allows resource-constrained clients to participate in training and inference. However, the distributed architecture exposes SL to backdoor attacks, where malicious clients can manipulate local datasets to alter the DNN's behavior. Existing defenses from other distributed frameworks like Federated Learning are not applicable, and there is a lack of effective backdoor defenses specifically designed for SL. We present SafeSplit, the first defense against client-side backdoor attacks in Split Learning (SL). SafeSplit enables the server to detect and filter out malicious client behavior by employing circular backward analysis after a client's training is completed, iteratively reverting to a trained checkpoint where the model under examination is found to be benign. It uses a two-fold analysis to identify client-induced changes and detect poisoned models. First, a static analysis in the frequency domain measures the differences in the layer's parameters at the server. Second, a dynamic analysis introduces a novel rotational distance metric that assesses the orientation shifts of the server's layer parameters during training. Our comprehensive evaluation across various data distributions, client counts, and attack scenarios demonstrates the high efficacy of this dual analysis in mitigating backdoor attacks while preserving model utility.

Alessandro Pegoraro, Carlotta Segna, Kavita Kumari et al.

Deep Learning (DL) models have become crucial in digital transformation, thus raising concerns about their intellectual property rights. Different watermarking techniques have been developed to protect Deep Neural Networks (DNNs) from IP infringement, creating a competitive field for DNN watermarking and removal methods. The predominant watermarking schemes use white-box techniques, which involve modifying weights by adding a unique signature to specific DNN layers. On the other hand, existing attacks on white-box watermarking usually require knowledge of the specific deployed watermarking scheme or access to the underlying data for further training and fine-tuning. We propose DeepEclipse, a novel and unified framework designed to remove white-box watermarks. We present obfuscation techniques that significantly differ from the existing white-box watermarking removal schemes. DeepEclipse can evade watermark detection without prior knowledge of the underlying watermarking scheme, additional data, or training and fine-tuning. Our evaluation reveals that DeepEclipse excels in breaking multiple white-box watermarking schemes, reducing watermark detection to random guessing while maintaining a similar model accuracy as the original one. Our framework showcases a promising solution to address the ongoing DNN watermark protection and removal challenges.

Nojan Sheybani, Alessandro Pegoraro, Jonathan Knauer et al.

Split Learning (SL) is a distributed learning approach that enables resource-constrained clients to collaboratively train deep neural networks (DNNs) by offloading most layers to a central server while keeping in- and output layers on the client-side. This setup enables SL to leverage server computation capacities without sharing data, making it highly effective in resource-constrained environments dealing with sensitive data. However, the distributed nature enables malicious clients to manipulate the training process. By sending poisoned intermediate gradients, they can inject backdoors into the shared DNN. Existing defenses are limited by often focusing on server-side protection and introducing additional overhead for the server. A significant challenge for client-side defenses is enforcing malicious clients to correctly execute the defense algorithm. We present ZORRO, a private, verifiable, and robust SL defense scheme. Through our novel design and application of interactive zero-knowledge proofs (ZKPs), clients prove their correct execution of a client-located defense algorithm, resulting in proofs of computational integrity attesting to the benign nature of locally trained DNN portions. Leveraging the frequency representation of model partitions enables ZORRO to conduct an in-depth inspection of the locally trained models in an untrusted environment, ensuring that each client forwards a benign checkpoint to its succeeding client. In our extensive evaluation, covering different model architectures as well as various attack strategies and data scenarios, we show ZORRO's effectiveness, as it reduces the attack success rate to less than 6\% while causing even for models storing \numprint{1000000} parameters on the client-side an overhead of less than 10 seconds.

Thien Duc Nguyen, Markus Miettinen, Alexandra Dmitrienko et al.

The COVID-19 pandemic has caused many countries to deploy novel digital contact tracing (DCT) systems to boost the efficiency of manual tracing of infection chains. In this paper, we systematically analyze DCT solutions and categorize them based on their design approaches and architectures. We analyze them with regard to effectiveness, security, privacy, and ethical aspects and compare prominent solutions with regard to these requirements. In particular, we discuss the shortcomings of the Google and Apple Exposure Notification API (GAEN) that is currently widely adopted all over the world. We find that the security and privacy of GAEN have considerable deficiencies as it can be compromised by severe, large-scale attacks. We also discuss other proposed approaches for contact tracing, including our proposal TRACECORONA, that are based on Diffie-Hellman (DH) key exchange and aim at tackling shortcomings of existing solutions. Our extensive analysis shows thatTRACECORONA fulfills the above security requirements better than deployed state-of-the-art approaches. We have implementedTRACECORONA, and its beta test version has been used by more than 2000 users without any major functional problems, demonstrating that there are no technical reasons requiring to make compromises with regard to the requirements of DCTapproaches.

Phillip Rieger, Thien Duc Nguyen, Markus Miettinen et al.

Federated Learning (FL) allows multiple clients to collaboratively train a Neural Network (NN) model on their private data without revealing the data. Recently, several targeted poisoning attacks against FL have been introduced. These attacks inject a backdoor into the resulting model that allows adversary-controlled inputs to be misclassified. Existing countermeasures against backdoor attacks are inefficient and often merely aim to exclude deviating models from the aggregation. However, this approach also removes benign models of clients with deviating data distributions, causing the aggregated model to perform poorly for such clients. To address this problem, we propose DeepSight, a novel model filtering approach for mitigating backdoor attacks. It is based on three novel techniques that allow to characterize the distribution of data used to train model updates and seek to measure fine-grained differences in the internal structure and outputs of NNs. Using these techniques, DeepSight can identify suspicious model updates. We also develop a scheme that can accurately cluster model updates. Combining the results of both components, DeepSight is able to identify and eliminate model clusters containing poisoned models with high attack impact. We also show that the backdoor contributions of possibly undetected poisoned models can be effectively mitigated with existing weight clipping-based defenses. We evaluate the performance and effectiveness of DeepSight and show that it can mitigate state-of-the-art backdoor attacks with a negligible impact on the model's performance on benign data.

Ghada Dessouky, Alexander Gruler, Pouya Mahmoody et al.

Shared cache resources in multi-core processors are vulnerable to cache side-channel attacks. Recently proposed defenses have their own caveats: Randomization-based defenses are vulnerable to the evolving attack algorithms besides relying on weak cryptographic primitives, because they do not fundamentally address the root cause for cache side-channel attacks. Cache partitioning defenses, on the other hand, provide the strict resource partitioning and effectively block all side-channel threats. However, they usually rely on way-based partitioning which is not fine-grained and cannot scale to support a larger number of protection domains, e.g., in trusted execution environment (TEE) security architectures, besides degrading performance and often resulting in cache underutilization. To overcome the shortcomings of both approaches, we present a novel and flexible set-associative cache partitioning design for TEE architectures, called Chunked-Cache. Chunked-Cache enables an execution context to "carve" out an exclusive configurable chunk of the cache if the execution requires side-channel resilience. If side-channel resilience is not required, mainstream cache resources are freely utilized. Hence, our solution addresses the security-performance trade-off practically by enabling selective and on-demand utilization of side-channel-resilient caches, while providing well-grounded future-proof security guarantees. We show that Chunked-Cache provides side-channel-resilient cache utilization for sensitive code execution, with small hardware overhead, while incurring no performance overhead on the OS. We also show that it outperforms conventional way-based cache partitioning by 43%, while scaling significantly better to support a larger number of protection domains.

Yanjiao Chen, Yijie Bai, Richard Mitev et al.

In the area of Internet of Things (IoT) voice assistants have become an important interface to operate smart speakers, smartphones, and even automobiles. To save power and protect user privacy, voice assistants send commands to the cloud only if a small set of pre-registered wake-up words are detected. However, voice assistants are shown to be vulnerable to the FakeWake phenomena, whereby they are inadvertently triggered by innocent-sounding fuzzy words. In this paper, we present a systematic investigation of the FakeWake phenomena from three aspects. To start with, we design the first fuzzy word generator to automatically and efficiently produce fuzzy words instead of searching through a swarm of audio materials. We manage to generate 965 fuzzy words covering 8 most popular English and Chinese smart speakers. To explain the causes underlying the FakeWake phenomena, we construct an interpretable tree-based decision model, which reveals phonetic features that contribute to false acceptance of fuzzy words by wake-up word detectors. Finally, we propose remedies to mitigate the effect of FakeWake. The results show that the strengthened models are not only resilient to fuzzy words but also achieve better overall performance on original training datasets.

Thien Duc Nguyen, Phillip Rieger, Huili Chen et al.

Federated Learning (FL) is a collaborative machine learning approach allowing participants to jointly train a model without having to share their private, potentially sensitive local datasets with others. Despite its benefits, FL is vulnerable to backdoor attacks, in which an adversary injects manipulated model updates into the model aggregation process so that the resulting model will provide targeted false predictions for specific adversary-chosen inputs. Proposed defenses against backdoor attacks based on detecting and filtering out malicious model updates consider only very specific and limited attacker models, whereas defenses based on differential privacy-inspired noise injection significantly deteriorate the benign performance of the aggregated model. To address these deficiencies, we introduce FLAME, a defense framework that estimates the sufficient amount of noise to be injected to ensure the elimination of backdoors while maintaining the model performance. To minimize the required amount of noise, FLAME uses a model clustering and weight clipping approach. Our evaluation of FLAME on several datasets stemming from application areas including image classification, word prediction, and IoT intrusion detection demonstrates that FLAME removes backdoors effectively with a negligible impact on the benign performance of the models. Furthermore, following the considerable attention that our research has received after its presentation at USENIX SEC 2022, FLAME has become the subject of numerous investigations proposing diverse attack methodologies in an attempt to circumvent it. As a response to these endeavors, we provide a comprehensive analysis of these attempts. Our findings show that these papers (e.g., 3DFed [36]) have not fully comprehended nor correctly employed the fundamental principles underlying FLAME, i.e., our defense mechanism effectively repels these attempted attacks.

Raad Bahmani, Ferdinand Brasser, Ghada Dessouky et al.

Security architectures providing Trusted Execution Environments (TEEs) have been an appealing research subject for a wide range of computer systems, from low-end embedded devices to powerful cloud servers. The goal of these architectures is to protect sensitive services in isolated execution contexts, called enclaves. Unfortunately, existing TEE solutions suffer from significant design shortcomings. First, they follow a one-size-fits-all approach offering only a single enclave type, however, different services need flexible enclaves that can adjust to their demands. Second, they cannot efficiently support emerging applications (e.g., Machine Learning as a Service), which require secure channels to peripherals (e.g., accelerators), or the computational power of multiple cores. Third, their protection against cache side-channel attacks is either an afterthought or impractical, i.e., no fine-grained mapping between cache resources and individual enclaves is provided. In this work, we propose CURE, the first security architecture, which tackles these design challenges by providing different types of enclaves: (i) sub-space enclaves provide vertical isolation at all execution privilege levels, (ii) user-space enclaves provide isolated execution to unprivileged applications, and (iii) self-contained enclaves allow isolated execution environments that span multiple privilege levels. Moreover, CURE enables the exclusive assignment of system resources, e.g., peripherals, CPU cores, or cache resources to single enclaves. CURE requires minimal hardware changes while significantly improving the state of the art of hardware-assisted security architectures. We implemented CURE on a RISC-V-based SoC and thoroughly evaluated our prototype in terms of hardware and performance overhead. CURE imposes a geometric mean performance overhead of 15.33% on standard benchmarks.

Shaza Zeitouni, Ghada Dessouky, Ahmad-Reza Sadeghi

In their continuous growth and penetration into new markets, Field Programmable Gate Arrays (FPGAs) have recently made their way into hardware acceleration of machine learning among other specialized compute-intensive services in cloud data centers, such as Amazon and Microsoft. To further maximize their utilization in the cloud, several academic works propose the spatial multi-tenant deployment model, where the FPGA fabric is simultaneously shared among mutually mistrusting clients. This is enabled by leveraging the partial reconfiguration property of FPGAs, which allows to split the FPGA fabric into several logically isolated regions and reconfigure the functionality of each region independently at runtime. In this paper, we survey industrial and academic deployment models of multi-tenant FPGAs in the cloud computing settings, and highlight their different adversary models and security guarantees, while shedding light on their fundamental shortcomings from a security standpoint. We further survey and classify existing academic works that demonstrate a new class of remotely exploitable physical attacks on multi-tenant FPGA devices, where these attacks are launched remotely by malicious clients sharing physical resources with victim users. Through investigating the problem of end-to-end multi-tenant FPGA deployment more comprehensively, we reveal how these attacks actually represent only one dimension of the problem, while various open security and privacy challenges remain unaddressed. We conclude with our insights and a call for future research to tackle these challenges.

Rosario Cammarota, Matthias Schunter, Anand Rajan et al.

In this work, we provide an industry research view for approaching the design, deployment, and operation of trustworthy Artificial Intelligence (AI) inference systems. Such systems provide customers with timely, informed, and customized inferences to aid their decision, while at the same time utilizing appropriate security protection mechanisms for AI models. Additionally, such systems should also use Privacy-Enhancing Technologies (PETs) to protect customers' data at any time. To approach the subject, we start by introducing current trends in AI inference systems. We continue by elaborating on the relationship between Intellectual Property (IP) and private data protection in such systems. Regarding the protection mechanisms, we survey the security and privacy building blocks instrumental in designing, building, deploying, and operating private AI inference systems. For example, we highlight opportunities and challenges in AI systems using trusted execution environments combined with more recent advances in cryptographic techniques to protect data in use. Finally, we outline areas of further development that require the global collective attention of industry, academia, and government researchers to sustain the operation of trustworthy AI inference systems.

Sebastian P. Bayerl, Tommaso Frassetto, Patrick Jauernig et al.

Performing machine learning tasks in mobile applications yields a challenging conflict of interest: highly sensitive client information (e.g., speech data) should remain private while also the intellectual property of service providers (e.g., model parameters) must be protected. Cryptographic techniques offer secure solutions for this, but have an unacceptable overhead and moreover require frequent network interaction. In this work, we design a practically efficient hardware-based solution. Specifically, we build Offline Model Guard (OMG) to enable privacy-preserving machine learning on the predominant mobile computing platform ARM - even in offline scenarios. By leveraging a trusted execution environment for strict hardware-enforced isolation from other system components, OMG guarantees privacy of client data, secrecy of provided models, and integrity of processing algorithms. Our prototype implementation on an ARM HiKey 960 development board performs privacy-preserving keyword recognition using TensorFlow Lite for Microcontrollers in real time.

Richard Mitev, Anna Pazii, Markus Miettinen et al.

Manufacturers of smart home Internet of Things (IoT) devices are increasingly adding voice assistant and audio monitoring features to a wide range of devices including smart speakers, televisions, thermostats, security systems, and doorbells. Consequently, many of these devices are equipped with microphones, raising significant privacy concerns: users may not always be aware of when audio recordings are sent to the cloud, or who may gain access to the recordings. In this paper, we present the LeakyPick architecture that enables the detection of the smart home devices that stream recorded audio to the Internet without the user's consent. Our proof-of-concept is a LeakyPick device that is placed in a user's smart home and periodically "probes" other devices in its environment and monitors the subsequent network traffic for statistical patterns that indicate audio transmission. Our prototype is built on a Raspberry Pi for less than USD40 and has a measurement accuracy of 94% in detecting audio transmissions for a collection of 8 devices with voice assistant capabilities. Furthermore, we used LeakyPick to identify 89 words that an Amazon Echo Dot misinterprets as its wake-word, resulting in unexpected audio transmission. LeakyPick provides a cost effective approach for regular consumers to monitor their homes for unexpected audio transmissions to the cloud.

Lars Baumgärtner, Alexandra Dmitrienko, Bernd Freisleben et al.

Google and Apple have jointly provided an API for exposure notification in order to implement decentralized contract tracing apps using Bluetooth Low Energy, the so-called "Google/Apple Proposal", which we abbreviate by "GAP". We demonstrate that in real-world scenarios the current GAP design is vulnerable to (i) profiling and possibly de-anonymizing infected persons, and (ii) relay-based wormhole attacks that basically can generate fake contacts with the potential of affecting the accuracy of an app-based contact tracing system. For both types of attack, we have built tools that can easily be used on mobile phones or Raspberry Pis (e.g., Bluetooth sniffers). The goal of our work is to perform a reality check towards possibly providing empirical real-world evidence for these two privacy and security risks. We hope that our findings provide valuable input for developing secure and privacy-preserving digital contact tracing systems.

Zijo Kenjar, Tommaso Frassetto, David Gens et al.

Fault-injection attacks have been proven in the past to be a reliable way of bypassing hardware-based security measures, such as cryptographic hashes, privilege and access permission enforcement, and trusted execution environments. However, traditional fault-injection attacks require physical presence, and hence, were often considered out of scope in many real-world adversary settings. In this paper we show this assumption may no longer be justified. We present V0LTpwn, a novel hardware-oriented but software-controlled attack that affects the integrity of computation in virtually any execution mode on modern x86 processors. To the best of our knowledge, this represents the first attack on x86 integrity from software. The key idea behind our attack is to undervolt a physical core to force non-recoverable hardware faults. Under a V0LTpwn attack, CPU instructions will continue to execute with erroneous results and without crashes, allowing for exploitation. In contrast to recently presented side-channel attacks that leverage vulnerable speculative execution, V0LTpwn is not limited to information disclosure, but allows adversaries to affect execution, and hence, effectively breaks the integrity goals of modern x86 platforms. In our detailed evaluation we successfully launch software-based attacks against Intel SGX enclaves from a privileged process to demonstrate that a V0LTpwn attack can successfully change the results of computations within enclave execution across multiple CPU revisions.

Tigist Abera, Ferdinand Brasser, Lachlan J. Gunn et al.

Autonomous collaborative networks of devices are rapidly emerging in numerous domains, such as self-driving cars, smart factories, critical infrastructure, and Internet of Things in general. Although autonomy and self-organization are highly desired properties, they increase vulnerability to attacks. Hence, autonomous networks need dependable mechanisms to detect malicious devices in order to prevent compromise of the entire network. However, current mechanisms to detect malicious devices either require a trusted central entity or scale poorly. In this paper, we present GrandDetAuto, the first scheme to identify malicious devices efficiently within large autonomous networks of collaborating entities. GrandDetAuto functions without relying on a central trusted entity, works reliably for very large networks of devices, and is adaptable to a wide range of application scenarios thanks to interchangeable components. Our scheme uses random elections to embed integrity validation schemes in distributed consensus, providing a solution supporting tens of thousands of devices. We implemented and evaluated a concrete instance of GrandDetAuto on a network of embedded devices and conducted large-scale network simulations with up to 100000 nodes. Our results show the effectiveness and efficiency of our scheme, revealing logarithmic growth in run-time and message complexity with increasing network size. Moreover, we provide an extensive evaluation of key parameters showing that GrandDetAuto is applicable to many scenarios with diverse requirements.

Ghada Dessouky, Tommaso Frassetto, Ahmad-Reza Sadeghi

Modern multi-core processors share cache resources for maximum cache utilization and performance gains. However, this leaves the cache vulnerable to side-channel attacks, where timing differences in shared cache behavior are exploited to infer information on the victim's execution patterns, ultimately leaking private information. The root cause for these attacks is mutually distrusting processes sharing cache entries and accessing them in a deterministic manner. Various defenses against cache side-channel attacks have been proposed. However, they either degrade performance significantly, impose impractical restrictions, or can only defeat certain classes of these attacks. More importantly, they assume that side-channel-resilient caches are required for the entire execution workload and do not allow to selectively enable the mitigation only for the security-critical portion of the workload. We present a generic mechanism for a flexible and soft partitioning of set-associative caches and propose a hybrid cache architecture, called HybCache. HybCache can be configured to selectively apply side-channel-resilient cache behavior only for isolated execution domains, while providing the non-isolated execution with conventional cache behavior, capacity and performance. An isolation domain can include one or more processes, specific portions of code, or a Trusted Execution Environment. We show that, with minimal hardware modifications and kernel support, HybCache can provide side-channel-resilient cache only for isolated execution with a performance overhead of 3.5-5%, while incurring no performance overhead for the remaining execution workload. We provide a simulator-based and hardware implementation of HybCache to evaluate the performance and area overheads, and show how it mitigates typical access-based and contention-based cache attacks.

Ebrahim M. Songhori, M. Sadegh Riazi, Siam U. Hussain et al.

We present ARM2GC, a novel secure computation framework based on Yao's Garbled Circuit (GC) protocol and the ARM processor. It allows users to develop privacy-preserving applications using standard high-level programming languages (e.g., C) and compile them using off-the-shelf ARM compilers (e.g., gcc-arm). The main enabler of this framework is the introduction of SkipGate, an algorithm that dynamically omits the communication and encryption cost of the gates whose outputs are independent of the private data. SkipGate greatly enhances the performance of ARM2GC by omitting costs of the gates associated with the instructions of the compiled binary, which is known by both parties involved in the computation. Our evaluation on benchmark functions demonstrates that ARM2GC not only outperforms the current GC frameworks that support high-level languages, it also achieves efficiency comparable to the best prior solutions based on hardware description languages. Moreover, in contrast to previous high-level frameworks with domain-specific languages and customized compilers, ARM2GC relies on standard ARM compiler which is rigorously verified and supports programs written in the standard syntax.

Sridhar Adepu, Ferdinand Brasser, Luis Garcia et al.

Cyber-physical control systems, such as industrial control systems (ICS), are increasingly targeted by cyberattacks. Such attacks can potentially cause tremendous damage, affect critical infrastructure or even jeopardize human life when the system does not behave as intended. Cyberattacks, however, are not new and decades of security research have developed plenty of solutions to thwart them. Unfortunately, many of these solutions cannot be easily applied to safety-critical cyber-physical systems. Further, the attack surface of ICS is quite different from what can be commonly assumed in classical IT systems. We present Scadman, a system with the goal to preserve the Control Behavior Integrity (CBI) of distributed cyber-physical systems. By observing the system-wide behavior, the correctness of individual controllers in the system can be verified. This allows Scadman to detect a wide range of attacks against controllers, like programmable logic controller (PLCs), including malware attacks, code-reuse and data-only attacks. We implemented and evaluated Scadman based on a real-world water treatment testbed for research and training on ICS security. Our results show that we can detect a wide range of attacks--including attacks that have previously been undetectable by typical state estimation techniques--while causing no false-positive warning for nominal threshold values.

Ghada Dessouky, David Gens, Patrick Haney et al.

In this paper, we take a deep dive into microarchitectural security from a hardware designer's perspective by reviewing the existing approaches to detect hardware vulnerabilities during the design phase. We show that a protection gap currently exists in practice that leaves chip designs vulnerable to software-based attacks. In particular, existing verification approaches fail to detect specific classes of vulnerabilities, which we call HardFails: these bugs evade detection by current verification techniques while being exploitable from software. We demonstrate such vulnerabilities in real-world SoCs using RISC-V to showcase and analyze concrete instantiations of HardFails. Patching these hardware bugs may not always be possible and can potentially result in a product recall. We base our findings on two extensive case studies: the recent Hack@DAC 2018 hardware security competition, where 54 independent teams of researchers competed world-wide over a period of 12 weeks to catch inserted security bugs in SoC RTL designs, and an in-depth systematic evaluation of state-of-the-art verification approaches. Our findings indicate that even combinations of techniques will miss high-impact bugs due to the large number of modules with complex interdependencies and fundamental limitations of current detection approaches. We also craft a real-world software attack that exploits one of the RTL bugs from Hack@DAC that evaded detection and discuss novel approaches to mitigate the growing problem of cross-layer bugs at design time.

Markus Miettinen, Paul C. van Oorschot, Ahmad-Reza Sadeghi

The emerging Internet of Things (IoT) drastically increases the number of connected devices in homes, workplaces and smart city infrastructures. This drives a need for means to not only ensure confidentiality of device-related communications, but for device configuration and management---ensuring that only legitimate devices are granted privileges to a local domain, that only authorized agents have access to the device and data it holds, and that software updates are authentic. The need to support device on-boarding, ongoing device management and control, and secure decommissioning dictates a suite of key management services for both access control to devices, and access by devices to wireless infrastructure and networked resources. We identify this core functionality, and argue for the recognition of efficient and reliable key management support---both within IoT devices, and by a unifying external management platform---as a baseline requirement for an IoT world. We present a framework architecture to facilitate secure, flexible and convenient device management in commodity IoT scenarios, and offer an illustrative set of protocols as a base solution---not to promote specific solution details, but to highlight baseline functionality to help domain owners oversee deployments of large numbers of independent multi-vendor IoT devices.

Abbas Acar, Hossein Fereidooni, Tigist Abera et al.

A myriad of IoT devices such as bulbs, switches, speakers in a smart home environment allow users to easily control the physical world around them and facilitate their living styles through the sensors already embedded in these devices. Sensor data contains a lot of sensitive information about the user and devices. However, an attacker inside or near a smart home environment can potentially exploit the innate wireless medium used by these devices to exfiltrate sensitive information from the encrypted payload (i.e., sensor data) about the users and their activities, invading user privacy. With this in mind,in this work, we introduce a novel multi-stage privacy attack against user privacy in a smart environment. It is realized utilizing state-of-the-art machine-learning approaches for detecting and identifying the types of IoT devices, their states, and ongoing user activities in a cascading style by only passively sniffing the network traffic from smart home devices and sensors. The attack effectively works on both encrypted and unencrypted communications. We evaluate the efficiency of the attack with real measurements from an extensive set of popular off-the-shelf smart home IoT devices utilizing a set of diverse network protocols like WiFi, ZigBee, and BLE. Our results show that an adversary passively sniffing the traffic can achieve very high accuracy (above 90%) in identifying the state and actions of targeted smart home devices and their users. To protect against this privacy leakage, we also propose a countermeasure based on generating spoofed traffic to hide the device states and demonstrate that it provides better protection than existing solutions.

N. Asokan, Thomas Nyman, Norrathep Rattanavipanon et al.

Secure firmware update is an important stage in the IoT device life-cycle. Prior techniques, designed for other computational settings, are not readily suitable for IoT devices, since they do not consider idiosyncrasies of a realistic large-scale IoT deployment. This motivates our design of ASSURED, a secure and scalable update framework for IoT. ASSURED includes all stakeholders in a typical IoT update ecosystem, while providing end-to-end security between manufacturers and devices. To demonstrate its feasibility and practicality, ASSURED is instantiated and experimentally evaluated on two commodity hardware platforms. Results show that ASSURED is considerably faster than current update mechanisms in realistic settings.

Thien Duc Nguyen, Samuel Marchal, Markus Miettinen et al.

IoT devices are increasingly deployed in daily life. Many of these devices are, however, vulnerable due to insecure design, implementation, and configuration. As a result, many networks already have vulnerable IoT devices that are easy to compromise. This has led to a new category of malware specifically targeting IoT devices. However, existing intrusion detection techniques are not effective in detecting compromised IoT devices given the massive scale of the problem in terms of the number of different types of devices and manufacturers involved. In this paper, we present DÏoT, an autonomous self-learning distributed system for detecting compromised IoT devices effectively. In contrast to prior work, DÏoT uses a novel self-learning approach to classify devices into device types and build normal communication profiles for each of these that can subsequently be used to detect anomalous deviations in communication patterns. DÏoT utilizes a federated learning approach for aggregating behavior profiles efficiently. To the best of our knowledge, it is the first system to employ a federated learning approach to anomaly-detection-based intrusion detection. Consequently, DÏoT can cope with emerging new and unknown attacks. We systematically and extensively evaluated more than 30 off-the-shelf IoT devices over a long term and show that DÏoT is highly effective (95.6% detection rate) and fast (~257 ms) at detecting devices compromised by, for instance, the infamous Mirai malware. DÏoT reported no false alarms when evaluated in a real-world smart home deployment setting.

Ferdinand Brasser, Srdjan Capkun, Alexandra Dmitrienko et al.

Recent research has demonstrated that Intel's SGX is vulnerable to software-based side-channel attacks. In a common attack, the adversary monitors CPU caches to infer secret-dependent data accesses patterns. Known defenses have major limitations, as they require either error-prone developer assistance, incur extremely high runtime overhead, or prevent only specific attacks. In this paper, we propose data location randomization as a novel defense against side-channel attacks that target data access patterns. Our goal is to break the link between the memory observations by the adversary and the actual data accesses by the victim. We design and implement a compiler-based tool called DR.SGX that instruments the enclave code, permuting data locations at fine granularity. To prevent correlation of repeated memory accesses we periodically re-randomize all enclave data. Our solution requires no developer assistance and strikes the balance between side-channel protection and performance based on an adjustable security parameter.

Hossein Fereidooni, Jiska Classen, Tom Spink et al.

Tens of millions of wearable fitness trackers are shipped yearly to consumers who routinely collect information about their exercising patterns. Smartphones push this health-related data to vendors' cloud platforms, enabling users to analyze summary statistics on-line and adjust their habits. Third-parties including health insurance providers now offer discounts and financial rewards in exchange for such private information and evidence of healthy lifestyles. Given the associated monetary value, the authenticity and correctness of the activity data collected becomes imperative. In this paper, we provide an in-depth security analysis of the operation of fitness trackers commercialized by Fitbit, the wearables market leader. We reveal an intricate security through obscurity approach implemented by the user activity synchronization protocol running on the devices we analyze. Although non-trivial to interpret, we reverse engineer the message semantics, demonstrate how falsified user activity reports can be injected, and argue that based on our discoveries, such attacks can be performed at scale to obtain financial gains. We further document a hardware attack vector that enables circumvention of the end-to-end protocol encryption present in the latest Fitbit firmware, leading to the spoofing of valid encrypted fitness data. Finally, we give guidelines for avoiding similar vulnerabilities in future system designs.

Yansong Gao, Said F. Al-Sarawi, Derek Abbott et al.

Physical unclonable functions (PUFs), as hardware security primitives, exploit manufacturing randomness to extract hardware instance-specific secrets. One of most popular structures is time-delay based Arbiter PUF attributing to large number of challenge response pairs (CRPs) yielded and its compact realization. However, modeling building attacks threaten most variants of APUFs that are usually employed for strong PUF-oriented application---lightweight authentication---without reliance on the securely stored digital secrets based standard cryptographic protocols. In this paper, we investigate a reconfigurable latent obfuscation technique endowed PUF construction, coined as OB-PUF, to maintain the security of elementary PUF CRPs enabled authentication where a CRP is never used more than once. The obfuscation---determined by said random patterns---conceals and distorts the relationship between challenge-response pairs capable of thwarting a model building adversary needing to know the exact relationship between challenges and responses. A bit further, the obfuscation is hidden and reconfigured on demand, in other words, the patterns are not only invisible but also act as one-time pads that are only employed once per authentication around and then discarded. As a consequence, the OB-PUF demonstrates significant resistance to the recent revealed powerful Evaluation Strategy (ES) based modeling attacks where the direct relationship between challenge and response is even not a must. The OB-PUF's uniqueness and reliability metrics are also systematically studied followed by formal authentication capability evaluations.

Ghada Dessouky, Shaza Zeitouni, Thomas Nyman et al.

Attacks targeting software on embedded systems are becoming increasingly prevalent. Remote attestation is a mechanism that allows establishing trust in embedded devices. However, existing attestation schemes are either static and cannot detect control-flow attacks, or require instrumentation of software incurring high performance overheads. To overcome these limitations, we present LO-FAT, the first practical hardware-based approach to control-flow attestation. By leveraging existing processor hardware features and commonly-used IP blocks, our approach enables efficient control-flow attestation without requiring software instrumentation. We show that our proof-of-concept implementation based on a RISC-V SoC incurs no processor stalls and requires reasonable area overhead.

Thomas Nyman, Ghada Dessouky, Shaza Zeitouni et al.

Widespread use of memory unsafe programming languages (e.g., C and C++) leaves many systems vulnerable to memory corruption attacks. A variety of defenses have been proposed to mitigate attacks that exploit memory errors to hijack the control flow of the code at run-time, e.g., (fine-grained) randomization or Control Flow Integrity. However, recent work on data-oriented programming (DOP) demonstrated highly expressive (Turing-complete) attacks, even in the presence of these state-of-the-art defenses. Although multiple real-world DOP attacks have been demonstrated, no efficient defenses are yet available. We propose run-time scope enforcement (RSE), a novel approach designed to efficiently mitigate all currently known DOP attacks by enforcing compile-time memory safety constraints (e.g., variable visibility rules) at run-time. We present HardScope, a proof-of-concept implementation of hardware-assisted RSE for the new RISC-V open instruction set architecture. We discuss our systematic empirical evaluation of HardScope which demonstrates that it can mitigate all currently known DOP attacks, and has a real-world performance overhead of 3.2% in embedded benchmarks.

Benny Fuhry, Raad Bahmani, Ferdinand Brasser et al.

Software-based approaches for search over encrypted data are still either challenged by lack of proper, low-leakage encryption or slow performance. Existing hardware-based approaches do not scale well due to hardware limitations and software designs that are not specifically tailored to the hardware architecture, and are rarely well analyzed for their security (e.g., the impact of side channels). Additionally, existing hardware-based solutions often have a large code footprint in the trusted environment susceptible to software compromises. In this paper we present HardIDX: a hardware-based approach, leveraging Intel's SGX, for search over encrypted data. It implements only the security critical core, i.e., the search functionality, in the trusted environment and resorts to untrusted software for the remainder. HardIDX is deployable as a highly performant encrypted database index: it is logarithmic in the size of the index and searches are performed within a few milliseconds rather than seconds. We formally model and prove the security of our scheme showing that its leakage is equivalent to the best known searchable encryption schemes. Our implementation has a very small code and memory footprint yet still scales to virtually unlimited search index sizes, i.e., size is limited only by the general - non-secure - hardware resources.

Dean Sullivan, Orlando Arias, David Gens et al.

Instruction set randomization (ISR) was initially proposed with the main goal of countering code-injection attacks. However, ISR seems to have lost its appeal since code-injection attacks became less attractive because protection mechanisms such as data execution prevention (DEP) as well as code-reuse attacks became more prevalent. In this paper, we show that ISR can be extended to also protect against code-reuse attacks while at the same time offering security guarantees similar to those of software diversity, control-flow integrity, and information hiding. We present Scylla, a scheme that deploys a new technique for in-place code encryption to hide the code layout of a randomized binary, and restricts the control flow to a benign execution path. This allows us to i) implicitly restrict control-flow targets to basic block entries without requiring the extraction of a control-flow graph, ii) achieve execution integrity within legitimate basic blocks, and iii) hide the underlying code layout under malicious read access to the program. Our analysis demonstrates that Scylla is capable of preventing state-of-the-art attacks such as just-in-time return-oriented programming (JIT-ROP) and crash-resistant oriented programming (CROP). We extensively evaluate our prototype implementation of Scylla and show feasible performance overhead. We also provide details on how this overhead can be significantly reduced with dedicated hardware support.

Ferdinand Brasser, Urs Müller, Alexandra Dmitrienko et al.

Side-channel information leakage is a known limitation of SGX. Researchers have demonstrated that secret-dependent information can be extracted from enclave execution through page-fault access patterns. Consequently, various recent research efforts are actively seeking countermeasures to SGX side-channel attacks. It is widely assumed that SGX may be vulnerable to other side channels, such as cache access pattern monitoring, as well. However, prior to our work, the practicality and the extent of such information leakage was not studied. In this paper we demonstrate that cache-based attacks are indeed a serious threat to the confidentiality of SGX-protected programs. Our goal was to design an attack that is hard to mitigate using known defenses, and therefore we mount our attack without interrupting enclave execution. This approach has major technical challenges, since the existing cache monitoring techniques experience significant noise if the victim process is not interrupted. We designed and implemented novel attack techniques to reduce this noise by leveraging the capabilities of the privileged adversary. Our attacks are able to recover confidential information from SGX enclaves, which we illustrate in two example cases: extraction of an entire RSA-2048 key during RSA decryption, and detection of specific human genome sequences during genomic indexing. We show that our attacks are more effective than previous cache attacks and harder to mitigate than previous SGX side-channel attacks.