Jean-Pierre Seifert

Niklas Pirnay, Ryan Sweke, Jens Eisert et al.

Density modelling is the task of learning an unknown probability density function from samples, and is one of the central problems of unsupervised machine learning. In this work, we show that there exists a density modelling problem for which fault-tolerant quantum computers can offer a super-polynomial advantage over classical learning algorithms, given standard cryptographic assumptions. Along the way, we provide a variety of additional results and insights, of potential interest for proving future distribution learning separations between quantum and classical learning algorithms. Specifically, we (a) provide an overview of the relationships between hardness results in supervised learning and distribution learning, and (b) show that any weak pseudo-random function can be used to construct a classically hard density modelling problem. The latter result opens up the possibility of proving quantum-classical separations for density modelling based on weaker assumptions than those necessary for pseudo-random functions.

Tommi Unruh, Bhargava Shastry, Malte Skoruppa et al.

The Web is replete with tutorial-style content on how to accomplish programming tasks. Unfortunately, even top-ranked tutorials suffer from severe security vulnerabilities, such as cross-site scripting (XSS), and SQL injection (SQLi). Assuming that these tutorials influence real-world software development, we hypothesize that code snippets from popular tutorials can be used to bootstrap vulnerability discovery at scale. To validate our hypothesis, we propose a semi-automated approach to find recurring vulnerabilities starting from a handful of top-ranked tutorials that contain vulnerable code snippets. We evaluate our approach by performing an analysis of tens of thousands of open-source web applications to check if vulnerabilities originating in the selected tutorials recur. Our analysis framework has been running on a standard PC, analyzed 64,415 PHP codebases hosted on GitHub thus far, and found a total of 117 vulnerabilities that have a strong syntactic similarity to vulnerable code snippets present in popular tutorials. In addition to shedding light on the anecdotal belief that programmers reuse web tutorial code in an ad hoc manner, our study finds disconcerting evidence of insufficiently reviewed tutorials compromising the security of open-source projects. Moreover, our findings testify to the feasibility of large-scale vulnerability discovery using poorly written tutorials as a starting point.

Niklas Pirnay, Anna Pappa, Jean-Pierre Seifert

Physical Unclonable Functions (PUFs) have been proposed as a way to identify and authenticate electronic devices. Recently, several ideas have been presented that aim to achieve the same for quantum devices. Some of these constructions apply single-qubit gates in order to provide a secure fingerprint of the quantum device. In this work, we formalize the class of Classical Readout Quantum PUFs (CR-QPUFs) using the statistical query (SQ) model and explicitly show insufficient security for CR-QPUFs based on single qubit rotation gates, when the adversary has SQ access to the CR-QPUF. We demonstrate how a malicious party can learn the CR-QPUF characteristics and forge the signature of a quantum device through a modelling attack using a simple regression of low-degree polynomials. The proposed modelling attack was successfully implemented in a real-world scenario on real IBM Q quantum machines. We thoroughly discuss the prospects and problems of CR-QPUFs where quantum device imperfections are used as a secure fingerprint.

Marcel Hinsche, Marios Ioannou, Alexander Nietner et al.

There is currently a large interest in understanding the potential advantages quantum devices can offer for probabilistic modelling. In this work we investigate, within two different oracle models, the probably approximately correct (PAC) learnability of quantum circuit Born machines, i.e., the output distributions of local quantum circuits. We first show a negative result, namely, that the output distributions of super-logarithmic depth Clifford circuits are not sample-efficiently learnable in the statistical query model, i.e., when given query access to empirical expectation values of bounded functions over the sample space. This immediately implies the hardness, for both quantum and classical algorithms, of learning from statistical queries the output distributions of local quantum circuits using any gate set which includes the Clifford group. As many practical generative modelling algorithms use statistical queries -- including those for training quantum circuit Born machines -- our result is broadly applicable and strongly limits the possibility of a meaningful quantum advantage for learning the output distributions of local quantum circuits. As a positive result, we show that in a more powerful oracle model, namely when directly given access to samples, the output distributions of local Clifford circuits are computationally efficiently PAC learnable by a classical learner. Our results are equally applicable to the problems of learning an algorithm for generating samples from the target distribution (generative modelling) and learning an algorithm for evaluating its probabilities (density modelling). They provide the first rigorous insights into the learnability of output distributions of local quantum circuits from the probabilistic modelling perspective.

Felicitas Hetzelt, Martin Radev, Robert Buhren et al.

Both AMD and Intel have presented technologies for confidential computing in cloud environments. The proposed solutions - AMD SEV (-ES, -SNP) and Intel TDX - protect Virtual Machines (VMs) against attacks from higher privileged layers through memory encryption and integrity protection. This model of computation draws a new trust boundary between virtual devices and the VM, which in so far lacks thorough examination. In this paper, we therefore present an analysis of the virtual device interface and discuss several attack vectors against a protected VM. Further, we develop and evaluate VIA, an automated analysis tool to detect cases of improper sanitization of input recieved via the virtual device interface. VIA improves upon existing approaches for the automated analysis of device interfaces in the following aspects: (i) support for virtualization relevant buses, (ii) efficient Direct Memory Access (DMA) support and (iii) performance. VIA builds upon the Linux Kernel Library and clang's libfuzzer to fuzz the communication between the driver and the device via MMIO, PIO, and DMA. An evaluation of VIA shows that it performs 570 executions per second on average and improves performance compared to existing approaches by an average factor of 2706. Using VIA, we analyzed 22 drivers in Linux 5.10.0-rc6, thereby uncovering 50 bugs and initiating multiple patches to the virtual device driver interface of Linux. To prove our findings criticality under the threat model of AMD SEV and Intel TDX, we showcase three exemplary attacks based on the bugs found. The attacks enable a malicious hypervisor to corrupt the memory and gain code execution in protected VMs with SEV-ES and are theoretically applicable to SEV-SNP and TDX.

Soundes Marzougui, Nils Wisiol, Patrick Gersch et al.

Due to the advancing development of quantum computers, practical attacks on conventional public-key cryptography may become feasible in the next few decades. To address this risk, post-quantum schemes that are secure against quantum attacks are being developed. Lattice-based algorithms are promising replacements for conventional schemes, with BLISS being one of the earliest post-quantum signature schemes in this family. However, required subroutines such as Gaussian sampling have been demonstrated to be a risk for the security of BLISS, since implementing Gaussian sampling both efficient and secure with respect to physical attacks is highly challenging. This paper presents three related power side-channel attacks on GALACTICS, the latest constant-time implementation of BLISS. All attacks are based on leakages we identified in the Gaussian sampling and signing algorithm of GALACTICS. To run the attack, a profiling phase on a device identical to the device under attack is required to train machine learning classifiers. In the attack phase, the leakages of GALACTICS enable the trained classifiers to predict sensitive internal information with high accuracy, paving the road for three different key recovery attacks. We demonstrate the leakages by running GALACTICS on a Cortex-M4 and provide proof-of-concept data and implementation for all our attacks.

Otto Bittner, Thilo Krachenfels, Andreas Galauner et al.

Voltage fault injection (FI) is a well-known attack technique that can be used to force faulty behavior in processors during their operation. Glitching the supply voltage can cause data value corruption, skip security checks, or enable protected code paths. At the same time, modern systems on a chip (SoCs) are used in security-critical applications, such as self-driving cars and autonomous machines. Since these embedded devices are often physically accessible by attackers, vendors must consider device tampering in their threat models. However, while the threat of voltage FI is known since the early 2000s, it seems as if vendors still forget to integrate countermeasures. This work shows how the entire boot security of an Nvidia SoC, used in Tesla's autopilot and Mercedes-Benz's infotainment system, can be circumvented using voltage FI. We uncover a hidden bootloader that is only available to the manufacturer for testing purposes and disabled by fuses in shipped products. We demonstrate how to re-enable this bootloader using FI to gain code execution with the highest privileges, enabling us to extract the bootloader's firmware and decryption keys used in later boot stages. Using a hardware implant, an adversary might misuse the hidden bootloader to bypass trusted code execution even during the system's regular operation.

Robert Buhren, Hans Niklas Jacob, Thilo Krachenfels et al.

AMD Secure Encrypted Virtualization (SEV) offers protection mechanisms for virtual machines in untrusted environments through memory and register encryption. To separate security-sensitive operations from software executing on the main x86 cores, SEV leverages the AMD Secure Processor (AMD-SP). This paper introduces a new approach to attack SEV-protected virtual machines (VMs) by targeting the AMD-SP. We present a voltage glitching attack that allows an attacker to execute custom payloads on the AMD-SPs of all microarchitectures that support SEV currently on the market (Zen 1, Zen 2, and Zen 3). The presented methods allow us to deploy a custom SEV firmware on the AMD-SP, which enables an adversary to decrypt a VM's memory. Furthermore, using our approach, we can extract endorsement keys of SEV-enabled CPUs, which allows us to fake attestation reports or to pose as a valid target for VM migration without requiring physical access to the target host. Moreover, we reverse-engineered the Versioned Chip Endorsement Key (VCEK) mechanism introduced with SEV Secure Nested Paging (SEV-SNP). The VCEK binds the endorsement keys to the firmware version of TCB components relevant for SEV. Building on the ability to extract the endorsement keys, we show how to derive valid VCEKs for arbitrary firmware versions. With our findings, we prove that SEV cannot adequately protect confidential data in cloud environments from insider attackers, such as rogue administrators, on currently available CPUs.

Thilo Krachenfels, Jean-Pierre Seifert, Shahin Tajik

The threat of hardware Trojans (HTs) and their detection is a widely studied field. While the effort for inserting a Trojan into an application-specific integrated circuit (ASIC) can be considered relatively high, especially when trusting the chip manufacturer, programmable hardware is vulnerable to Trojan insertion even after the product has been shipped or during usage. At the same time, detecting dormant HTs with small or zero-overhead triggers and payloads on these platforms is still a challenging task, as the Trojan might not get activated during the chip verification using logical testing or physical measurements. In this work, we present a novel Trojan detection approach based on a technique known from integrated circuit (IC) failure analysis, capable of detecting virtually all classes of dormant Trojans. Using laser logic state imaging (LLSI), we show how supply voltage modulations can awaken inactive Trojans, making them detectable using laser voltage imaging techniques. Therefore, our technique does not require triggering the Trojan. To support our claims, we present three case studies on 28 and 20 SRAM- and flash-based field-programmable gate arrays (FPGAs). We demonstrate how to detect with high confidence small changes in sequential and combinatorial logic as well as in the routing configuration of FPGAs in a non-invasive manner. Finally, we discuss the practical applicability of our approach on dormant analog Trojans in ASICs.

Thilo Krachenfels, Tuba Kiyan, Shahin Tajik et al.

The security of modern electronic devices relies on secret keys stored on secure hardware modules as the root-of-trust (RoT). Extracting those keys would break the security of the entire system. As shown before, sophisticated side-channel analysis (SCA) attacks, using chip failure analysis (FA) techniques, can extract data from on-chip memory cells. However, since the chip's layout is unknown to the adversary in practice, secret key localization and reverse engineering are onerous tasks. Consequently, hardware vendors commonly believe that the ever-growing physical complexity of the integrated circuit (IC) designs can be a natural barrier against potential adversaries. In this work, we present a novel approach that can extract the secret key without any knowledge of the IC's layout, and independent from the employed memory technology as key storage. We automate the -- traditionally very labor-intensive -- reverse engineering and data extraction process. To that end, we demonstrate that black-box measurements captured using laser-assisted SCA techniques from a training device with known key can be used to profile the device for a later key prediction on other victim devices with unknown keys. To showcase the potential of our approach, we target keys on three different hardware platforms, which are utilized as RoT in different products.

Thilo Krachenfels, Fatemeh Ganji, Amir Moradi et al.

Due to its sound theoretical basis and practical efficiency, masking has become the most prominent countermeasure to protect cryptographic implementations against physical side-channel attacks (SCAs). The core idea of masking is to randomly split every sensitive intermediate variable during computation into at least t+1 shares, where t denotes the maximum number of shares that are allowed to be observed by an adversary without learning any sensitive information. In other words, it is assumed that the adversary is bounded either by the possessed number of probes (e.g., microprobe needles) or by the order of statistical analyses while conducting higher-order SCA attacks (e.g., differential power analysis). Such bounded models are employed to prove the SCA security of the corresponding implementations. Consequently, it is believed that given a sufficiently large number of shares, the vast majority of known SCA attacks are mitigated. In this work, we present a novel laser-assisted SCA technique, called Laser Logic State Imaging (LLSI), which offers an unlimited number of contactless probes, and therefore, violates the probing security model assumption. This technique enables us to take snapshots of hardware implementations, i.e., extract the logical state of all registers at any arbitrary clock cycle with a single measurement. To validate this, we mount our attack on masked AES hardware implementations and practically demonstrate the extraction of the full-length key in two different scenarios. First, we assume that the location of the registers (key and/or state) is known, and hence, their content can be directly read by a single snapshot. Second, we consider an implementation with unknown register locations, where we make use of multiple snapshots and a SAT solver to reveal the secrets.

Ryan Sweke, Jean-Pierre Seifert, Dominik Hangleiter et al.

Here we study the comparative power of classical and quantum learners for generative modelling within the Probably Approximately Correct (PAC) framework. More specifically we consider the following task: Given samples from some unknown discrete probability distribution, output with high probability an efficient algorithm for generating new samples from a good approximation of the original distribution. Our primary result is the explicit construction of a class of discrete probability distributions which, under the decisional Diffie-Hellman assumption, is provably not efficiently PAC learnable by a classical generative modelling algorithm, but for which we construct an efficient quantum learner. This class of distributions therefore provides a concrete example of a generative modelling problem for which quantum learners exhibit a provable advantage over classical learning algorithms. In addition, we discuss techniques for proving classical generative modelling hardness results, as well as the relationship between the PAC learnability of Boolean functions and the PAC learnability of discrete probability distributions.

Thilo Krachenfels, Heiko Lohrke, Jean-Pierre Seifert et al.

Recent attacks using thermal laser stimulation (TLS) have shown that it is possible to extract cryptographic keys from the battery-backed memory on state-of-the-art field-programmable gate arrays (FPGAs). However, the professional failure analysis microscopes usually employed for these attacks cost in the order of 500k to 1M dollars. In this work, we evaluate the use of a cheaper commercial laser fault injection station retrofitted with a suitable amplifier and light source to enable TLS. We demonstrate that TLS attacks are possible at a hardware cost of around 100k dollars. This constitutes a reduction of the resources required by the attacker by a factor of at least five. We showcase two actual attacks: data extraction from the SRAM memory of a low-power microcontroller and decryption key extraction from a 20-nm technology FPGA device. The strengths and weaknesses of our low-cost approach are then discussed in comparison with the conventional failure analysis equipment approach. In general, this work demonstrates that TLS backside attacks are available at a much lower cost than previously expected.

Robert Buhren, Christian Werling, Jean-Pierre Seifert

Customers of cloud services have to trust the cloud providers, as they control the building blocks that form the cloud. This includes the hypervisor enabling the sharing of a single hardware platform among multiple tenants. AMD Secure Encrypted Virtualization (SEV) claims a new level of protection in cloud scenarios. AMD SEV encrypts the main memory of virtual machines with VM-specific keys, thereby denying the higher-privileged hypervisor access to a guest's memory. To enable the cloud customer to verify the correct deployment of his virtual machine, SEV additionally introduces a remote attestation protocol.This paper analyzes the firmware components that implement the SEV remote attestation protocol on the current AMD Epyc Naples CPU series. We demonstrate that it is possible to extract critical CPU-specific keys that are fundamental for the security of the remote attestation protocol.Building on the extracted keys, we propose attacks that allow a malicious cloud provider a complete circumvention of the SEV protection mechanisms. Although the underlying firmware issues were already fixed by AMD, we show that the current series of AMD Epyc CPUs, i.e., the Naples series, does not prevent the installation of previous firmware versions. We show that the severity of our proposed attacks is very high as no purely software-based mitigations are possible. This effectively renders the SEV technology on current AMD Epyc CPUs useless when confronted with an untrusted cloud provider. To overcome these issues, we also propose robust changes to the SEV design that allow future generations of the SEV technology to mitigate the proposed attacks.

Soroush M. Sohi, Jean-Pierre Seifert, Fatemeh Ganji

Security of information passing through the Internet is threatened by today's most advanced malware ranging from orchestrated botnets to simpler polymorphic worms. These threats, as examples of zero-day attacks, are able to change their behavior several times in the early phases of their existence to bypass the network intrusion detection systems (NIDS). In fact, even well-designed, and frequently-updated signature-based NIDS cannot detect the zero-day treats due to the lack of an adequate signature database, adaptive to intelligent attacks on the Internet. More importantly, having an NIDS, it should be tested on malicious traffic dataset that not only represents known attacks, but also can to some extent reflect the characteristics of unknown, zero-day attacks. Generating such traffic is identified in the literature as one of the main obstacles for evaluating the effectiveness of NIDS. To address these issues, we introduce RNNIDS that applies Recurrent Neural Networks (RNNs) to find complex patterns in attacks and generate similar ones. In this regard, for the first time, we demonstrate that RNNs are helpful to generate new, unseen mutants of attacks as well as synthetic signatures from the most advanced malware to improve the intrusion detection rate. Besides, to further enhance the design of an NIDS, RNNs can be employed to generate malicious datasets containing, e.g., unseen mutants of a malware. To evaluate the feasibility of our approaches, we conduct extensive experiments by incorporating publicly available datasets, where we show a considerable improvement in the detection rate of an off-the-shelf NIDS (up to 16.67%).

Bhargava Shastry, Federico Maggi, Fabian Yamaguchi et al.

Taint-style vulnerabilities comprise a majority of fuzzer discovered program faults. These vulnerabilities usually manifest as memory access violations caused by tainted program input. Although fuzzers have helped uncover a majority of taint-style vulnerabilities in software to date, they are limited by (i) extent of test coverage; and (ii) the availability of fuzzable test cases. Therefore, fuzzing alone cannot provide a high assurance that all taint-style vulnerabilities have been uncovered. In this paper, we use static template matching to find recurrences of fuzzer-discovered vulnerabilities. To compensate for the inherent incompleteness of template matching, we implement a simple yet effective match-ranking algorithm that uses test coverage data to focus attention on those matches that comprise untested code. We prototype our approach using the Clang/LLVM compiler toolchain and use it in conjunction with afl-fuzz, a modern coverage-guided fuzzer. Using a case study carried out on the Open vSwitch codebase, we show that our prototype uncovers corner cases in modules that lack a fuzzable test harness. Our work demonstrates that static analysis can effectively complement fuzz testing, and is a useful addition to the security assessment tool-set. Furthermore, our techniques hold promise for increasing the effectiveness of program analysis and testing, and serve as a building block for a hybrid vulnerability discovery framework.

Robert Buhren, Shay Gueron, Jan Nordholz et al.

Adversaries with physical access to a target platform can perform cold boot or DMA attacks to extract sensitive data from the RAM. In response, several main-memory encryption schemes have been proposed to prevent such attacks. Also hardware vendors have acknowledged the threat and already announced respective hardware extensions. Intel's SGX and AMD's SME will provide means to encrypt parts of the RAM to protect security-relevant assets that reside there. Encrypting the RAM will protect the user's content against passive eavesdropping. However, the level of protection it provides in scenarios that involve an adversary who is not only able to read from RAM but can also change content in RAM is less clear. Obviously, encryption offers some protection against such an "active" adversary: from the ciphertext the adversary cannot see what value is changed in the plaintext, nor predict the system behaviour based on the changes. But is this enough to prevent an active adversary from performing malicious tasks? This paper addresses the open research question whether encryption alone is a dependable protection mechanism in practice when considering an active adversary. To this end, we first build a software based memory encryption solution on a desktop system which mimics AMD's SME. Subsequently, we demonstrate a proof-of-concept fault attack on this system, by which we are able to extract the private RSA key of a GnuPG user. Our work suggests that transparent memory encryption is not enough to prevent active attacks.

Altaf Shaik, Ravishankar Borgaonkar, N. Asokan et al.

Mobile communication systems now constitute an essential part of life throughout the world. Fourth generation "Long Term Evolution" (LTE) mobile communication networks are being deployed. The LTE suite of specifications is considered to be significantly better than its predecessors not only in terms of functionality but also with respect to security and privacy for subscribers. We carefully analyzed LTE access network protocol specifications and uncovered several vulnerabilities. Using commercial LTE mobile devices in real LTE networks, we demonstrate inexpensive, and practical attacks exploiting these vulnerabilities. Our first class of attacks consists of three different ways of making an LTE device leak its location: A semi-passive attacker can locate an LTE device within a 2 sq.km area within a city whereas an active attacker can precisely locate an LTE device using GPS co-ordinates or trilateration via cell-tower signal strength information. Our second class of attacks can persistently deny some or all services to a target LTE device. To the best of our knowledge, our work constitutes the first publicly reported practical attacks against LTE access network protocols. We present several countermeasures to resist our specific attacks. We also discuss possible trade-offs that may explain why these vulnerabilities exist and recommend that safety margins introduced into future specifications to address such trade-offs should incorporate greater agility to accommodate subsequent changes in the trade-off equilibrium.

Bhargava Shastry, Fabian Yamaguchi, Konrad Rieck et al.

Eliminating vulnerabilities from low-level code is vital for securing software. Static analysis is a promising approach for discovering vulnerabilities since it can provide developers early feedback on the code they write. But, it presents multiple challenges not the least of which is understanding what makes a bug exploitable and conveying this information to the developer. In this paper, we present the design and implementation of a practical vulnerability assessment framework, called Melange. Melange performs data and control flow analysis to diagnose potential security bugs, and outputs well-formatted bug reports that help developers understand and fix security bugs. Based on the intuition that real-world vulnerabilities manifest themselves across multiple parts of a program, Melange performs both local and global analyses. To scale up to large programs, global analysis is demand-driven. Our prototype detects multiple vulnerability classes in C and C++ code including type confusion, and garbage memory reads. We have evaluated Melange extensively. Our case studies show that Melange scales up to large codebases such as Chromium, is easy-to-use, and most importantly, capable of discovering vulnerabilities in real-world code. Our findings indicate that static analysis is a viable reinforcement to the software testing tool set.

Daniel Defreez, Bhargava Shastry, Hao Chen et al.

With Firefox OS, Mozilla is making a serious push for an HTML5-based mobile platform. In order to assuage security concerns over providing hardware access to web applications, Mozilla has introduced a number of mechanisms that make the security landscape of Firefox OS distinct from both the desktop web and other mobile operating systems. From an application security perspective, the two most significant of these mechanisms are the the introduction of a default Content Security Policy and code review in the market. This paper describes how lightweight static analysis can augment these mechanisms to find vulnerabilities which have otherwise been missed. We provide examples of privileged applications in the market that contain vulnerabilities that can be automatically detected. In addition to these findings, we show some of the challenges that occur when desktop software is repurposed for a mobile operating system. In particular, we argue that the caching of certificate overrides across applications--a known problem in Firefox OS--generates a counter-intuitive user experience that detracts from the security of the system.