Martin Strohmeier

Christian Schroeder de Witt, Samuel Sokota, J. Zico Kolter et al.

Steganography is the practice of encoding secret information into innocuous content in such a manner that an adversarial third party would not realize that there is hidden meaning. While this problem has classically been studied in security literature, recent advances in generative models have led to a shared interest among security and machine learning researchers in developing scalable steganography techniques. In this work, we show that a steganography procedure is perfectly secure under Cachin (1998)'s information-theoretic model of steganography if and only if it is induced by a coupling. Furthermore, we show that, among perfectly secure procedures, a procedure maximizes information throughput if and only if it is induced by a minimum entropy coupling. These insights yield what are, to the best of our knowledge, the first steganography algorithms to achieve perfect security guarantees for arbitrary covertext distributions. To provide empirical validation, we compare a minimum entropy coupling-based approach to three modern baselines -- arithmetic coding, Meteor, and adaptive dynamic grouping -- using GPT-2, WaveRNN, and Image Transformer as communication channels. We find that the minimum entropy coupling-based approach achieves superior encoding efficiency, despite its stronger security constraints. In aggregate, these results suggest that it may be natural to view information-theoretic steganography through the lens of minimum entropy coupling.

Eric Jedermann, Piotr Kulpinski, Martin Strohmeier et al.

The Iridium Low Earth Orbit (LEO) satellite constellation remains a unique provider of global communications for critical industries, governments, and private users, serving over 2.5 million active subscribers despite recent market competition. In contrast to terrestrial wireless standards such as 3GPP, Iridium protocol specifications are proprietary and have not undergone rigorous, public, and systematic security evaluation. In this work, we present the first comprehensive security analysis of Iridium authentication and radio link protocols. We reverse engineer Iridium SIM-based authentication mechanism and demonstrate that the secret key can be extracted from the SIM card, enabling full device cloning and impersonation attacks. Leveraging a month-long dataset of Iridium up- and downlink satellite traffic, we further show that nearly all signaling and radio communication protocols currently in use lack encryption, resulting in the exposure of sensitive information in cleartext over the air such as login credentials and large volumes of personal data. Finally, we develop custom software-defined radio (SDR) tools to carry out spoofing and jamming attacks, revealing that modestly equipped adversaries can inject falsified messages or disrupt the Iridium service locally due to the absence of source authentication. Our findings uncover systemic vulnerabilities in the Iridium radio link and highlight the urgent need for users of critical applications to transition to more secure communication radio links.

James Pavur, Martin Strohmeier, Vincent Lenders et al.

Satellite broadband services are critical infrastructures enabling advanced technologies to function in the most remote regions of the globe. However, status-quo services are often unencrypted by default and vulnerable to eavesdropping attacks. In this paper, we challenge the historical perception that over-the-air security must trade off with TCP performance in high-latency satellite networks due to the deep-packet inspection requirements of Performance Enhancing Proxies (PEPs). After considering why prior work in this area has failed to find wide adoption, we present an open-source encrypted-by-default PEP - QPEP - which seeks to address these issues. QPEP is built around the open QUIC standard and designed so individual customers may adopt it without ISP involvement. QPEP's performance is assessed through simulations in a replicable docker-based testbed. Across many benchmarks and network conditions, QPEP is found to avoid the perceived security-encryption trade-off in PEP design. Compared to unencrypted PEP implementations, QPEP reduces average page load times by more than 30% while also offering over-the-air privacy. Compared to the traditional VPN encryption available to customers today, QPEP more than halves average page load times. Together, these experiments lead to the conclusion that QPEP represents a promising new approach to protecting modern satellite broadband connections.

Martin Strohmeier, Matthew Smith, Vincent Lenders et al.

In recent years, air traffic communication data has become easy to access, enabling novel research in many fields. Exploiting this new data source, a wide range of applications have emerged, from weather forecasting to stock market prediction, or the collection of information about military and government movements. Typically these applications require knowledge about the metadata of the aircraft, specifically its operator and the aircraft category. armasuisse Science + Technology, the R\&D agency for the Swiss Armed Forces, has been developing Classi-Fly, a novel approach to obtain metadata about aircraft based on their movement patterns. We validate Classi-Fly using several hundred thousand flights collected through open source means, in conjunction with ground truth from publicly available aircraft registries containing more than two million aircraft. Classi-Fly obtains the correct aircraft category with an accuracy of over 88%, demonstrating that it can improve the meta data necessary for applications working with air traffic communication. Finally, we show that it is feasible to automatically detect specific flights such as police and surveillance missions.

Dario Pasquini, Martin Strohmeier, Carmela Troncoso

We introduce a new family of prompt injection attacks, termed Neural Exec. Unlike known attacks that rely on handcrafted strings (e.g., "Ignore previous instructions and..."), we show that it is possible to conceptualize the creation of execution triggers as a differentiable search problem and use learning-based methods to autonomously generate them. Our results demonstrate that a motivated adversary can forge triggers that are not only drastically more effective than current handcrafted ones but also exhibit inherent flexibility in shape, properties, and functionality. In this direction, we show that an attacker can design and generate Neural Execs capable of persisting through multi-stage preprocessing pipelines, such as in the case of Retrieval-Augmented Generation (RAG)-based applications. More critically, our findings show that attackers can produce triggers that deviate markedly in form and shape from any known attack, sidestepping existing blacklist-based detection and sanitation approaches.

Sumeet Ramesh Motwani, Mikhail Baranchuk, Martin Strohmeier et al.

Recent capability increases in large language models (LLMs) open up applications in which groups of communicating generative AI agents solve joint tasks. This poses privacy and security challenges concerning the unauthorised sharing of information, or other unwanted forms of agent coordination. Modern steganographic techniques could render such dynamics hard to detect. In this paper, we comprehensively formalise the problem of secret collusion in systems of generative AI agents by drawing on relevant concepts from both AI and security literature. We study incentives for the use of steganography, and propose a variety of mitigation measures. Our investigations result in a model evaluation framework that systematically tests capabilities required for various forms of secret collusion. We provide extensive empirical results across a range of contemporary LLMs. While the steganographic capabilities of current models remain limited, GPT-4 displays a capability jump suggesting the need for continuous monitoring of steganographic frontier model capabilities. We conclude by laying out a comprehensive research program to mitigate future risks of collusion between generative AI models.

Sebastian Köhler, Richard Baker, Martin Strohmeier et al.

We present a novel attack against the Combined Charging System, one of the most widely used DC rapid charging technologies for electric vehicles (EVs). Our attack, Brokenwire, interrupts necessary control communication between the vehicle and charger, causing charging sessions to abort. The attack requires only temporary physical proximity and can be conducted wirelessly from a distance, allowing individual vehicles or entire fleets to be disrupted stealthily and simultaneously. In addition, it can be mounted with off-the-shelf radio hardware and minimal technical knowledge. By exploiting CSMA/CA behavior, only a very weak signal needs to be induced into the victim to disrupt communication - exceeding the effectiveness of broadband noise jamming by three orders of magnitude. The exploited behavior is a required part of the HomePlug Green PHY, DIN 70121 & ISO 15118 standards and all known implementations exhibit it. We first study the attack in a controlled testbed and then demonstrate it against eight vehicles and 20 chargers in real deployments. We find the attack to be successful in the real world, at ranges up to 47 m, for a power budget of less than 1 W. We further show that the attack can work between the floors of a building (e.g., multi-story parking), through perimeter fences, and from `drive-by' attacks. We present a heuristic model to estimate the number of vehicles that can be attacked simultaneously for a given output power. Brokenwire has immediate implications for a substantial proportion of the around 12 million battery EVs on the roads worldwide - and profound effects on the new wave of electrification for vehicle fleets, both for private enterprise and crucial public services, as well as electric buses, trucks and small ships. As such, we conducted a disclosure to the industry and discussed a range of mitigation techniques that could be deployed to limit the impact.

Christian Schroeder de Witt, Yongchao Huang, Philip H. S. Torr et al.

Cyber attacks are increasing in volume, frequency, and complexity. In response, the security community is looking toward fully automating cyber defense systems using machine learning. However, so far the resultant effects on the coevolutionary dynamics of attackers and defenders have not been examined. In this whitepaper, we hypothesise that increased automation on both sides will accelerate the coevolutionary cycle, thus begging the question of whether there are any resultant fixed points, and how they are characterised. Working within the threat model of Locked Shields, Europe's largest cyberdefense exercise, we study blackbox adversarial attacks on network classifiers. Given already existing attack capabilities, we question the utility of optimal evasion attack frameworks based on minimal evasion distances. Instead, we suggest a novel reinforcement learning setting that can be used to efficiently generate arbitrary adversarial perturbations. We then argue that attacker-defender fixed points are themselves general-sum games with complex phase transitions, and introduce a temporally extended multi-agent reinforcement learning framework in which the resultant dynamics can be studied. We hypothesise that one plausible fixed point of AI-NIDS may be a scenario where the defense strategy relies heavily on whitelisted feature flow subspaces. Finally, we demonstrate that a continual learning approach is required to study attacker-defender dynamics in temporally extended general-sum games.

Samuel Sokota, Christian Schroeder de Witt, Maximilian Igl et al.

We consider the problem of communicating exogenous information by means of Markov decision process trajectories. This setting, which we call a Markov coding game (MCG), generalizes both source coding and a large class of referential games. MCGs also isolate a problem that is important in decentralized control settings in which cheap-talk is not available -- namely, they require balancing communication with the associated cost of communicating. We contribute a theoretically grounded approach to MCGs based on maximum entropy reinforcement learning and minimum entropy coupling that we call MEME. Due to recent breakthroughs in approximation algorithms for minimum entropy coupling, MEME is not merely a theoretical algorithm, but can be applied to practical settings. Empirically, we show both that MEME is able to outperform a strong baseline on small MCGs and that MEME is able to achieve strong performance on extremely large MCGs. To the latter point, we demonstrate that MEME is able to losslessly communicate binary images via trajectories of Cartpole and Pong, while simultaneously achieving the maximal or near maximal expected returns, and that it is even capable of performing well in the presence of actuator noise.

Matthew Smith, Martin Strohmeier, Vincent Lenders et al.

Airborne collision avoidance systems provide an onboard safety net should normal air traffic control procedures fail to keep aircraft separated. These systems are widely deployed and have been constantly refined over the past three decades, usually in response to near misses or mid-air collisions. Recent years have seen security research increasingly focus on aviation, identifying that key wireless links---some of which are used in collision avoidance---are vulnerable to attack. In this paper, we go one step further to understand whether an attacker can remotely trigger false collision avoidance alarms. Primarily considering the next-generation Airborne Collision Avoidance System X (ACAS X), we adopt a modelling approach to extract attacker constraints from technical standards before simulating collision avoidance attacks against standardized ACAS X code. We find that in 44% of cases, an attacker can successfully trigger a collision avoidance alert which on average results in a 590 ft altitude deviation; when the aircraft is at lower altitudes, this success rate rises considerably to 79%. Furthermore, we show how our simulation approach can be used to help defend against attacks by identifying where attackers are most likely to be successful.

Giulio Lovisotto, Henry Turner, Ivo Sluganovic et al.

Research into adversarial examples (AE) has developed rapidly, yet static adversarial patches are still the main technique for conducting attacks in the real world, despite being obvious, semi-permanent and unmodifiable once deployed. In this paper, we propose Short-Lived Adversarial Perturbations (SLAP), a novel technique that allows adversaries to realize physically robust real-world AE by using a light projector. Attackers can project a specifically crafted adversarial perturbation onto a real-world object, transforming it into an AE. This allows the adversary greater control over the attack compared to adversarial patches: (i) projections can be dynamically turned on and off or modified at will, (ii) projections do not suffer from the locality constraint imposed by patches, making them harder to detect. We study the feasibility of SLAP in the self-driving scenario, targeting both object detector and traffic sign recognition tasks, focusing on the detection of stop signs. We conduct experiments in a variety of ambient light conditions, including outdoors, showing how in non-bright settings the proposed method generates AE that are extremely robust, causing misclassifications on state-of-the-art networks with up to 99% success rate for a variety of angles and distances. We also demostrate that SLAP-generated AE do not present detectable behaviours seen in adversarial patches and therefore bypass SentiNet, a physical AE detection method. We evaluate other defences including an adaptive defender using adversarial learning which is able to thwart the attack effectiveness up to 80% even in favourable attacker conditions.

Matthew Smith, Martin Strohmeier, Jon Harman et al.

Many wireless communications systems found in aircraft lack standard security mechanisms, leaving them fundamentally vulnerable to attack. With affordable software-defined radios available, a novel threat has emerged, allowing a wide range of attackers to easily interfere with wireless avionic systems. Whilst these vulnerabilities are known, concrete attacks that exploit them are still novel and not yet well understood. This is true in particular with regards to their kinetic impact on the handling of the attacked aircraft and consequently its safety. To investigate this, we invited 30 Airbus A320 type-rated pilots to fly simulator scenarios in which they were subjected to attacks on their avionics. We implement and analyse novel wireless attacks on three safety-related systems: Traffic Collision Avoidance System (TCAS), Ground Proximity Warning System (GPWS) and the Instrument Landing System (ILS). We found that all three analysed attack scenarios created significant control impact and cost of disruption through turnarounds, avoidance manoeuvres, and diversions. They further increased workload, distrust in the affected system, and in 38% of cases caused the attacked safety system to be switched off entirely. All pilots felt the scenarios were useful, with 93.3% feeling that simulator training for wireless attacks could be valuable.

Matthew Smith, Daniel Moser, Martin Strohmeier et al.

The manner in which Aircraft Communications, Addressing and Reporting System (ACARS) is being used has significantly changed over time. Whilst originally used by commercial airliners to track their flights and provide automated timekeeping on crew, today it serves as a multi-purpose air-ground data link for many aviation stakeholders including private jet owners, state actors and military. Since ACARS messages are still mostly sent in the clear over a wireless channel, any sensitive information sent with ACARS can potentially lead to a privacy breach for users. Naturally, different stakeholders consider different types of data sensitive. In this paper we propose a privacy framework matching aviation stakeholders to a range of sensitive information types and assess the impact for each. Based on more than one million ACARS messages, collected over several months, we then demonstrate that current ACARS usage systematically breaches privacy for all stakeholder groups. We further support our findings with a number of cases of significant privacy issues for each group and analyze the impact of such leaks. While it is well-known that ACARS messages are susceptible to eavesdropping attacks, this work is the first to quantify the extent and impact of privacy leakage in the real world for the relevant aviation stakeholders.

Martin Strohmeier, Matthias Schäfer, Rui Pinheiro et al.

More than a dozen wireless technologies are used by air traffic communication systems during different flight phases. From a conceptual perspective, all of them are insecure as security was never part of their design. Recent contributions from academic and hacking communities have exploited this inherent vulnerability to demonstrate attacks on some of these technologies. However, not all of these contributions have resonated widely within aviation circles. At the same time, the security community lacks certain aviation domain knowledge, preventing aviation authorities from giving credence to their findings. In this paper, we aim to reconcile the view of the security community and the perspective of aviation professionals concerning the safety of air traffic communication technologies. To achieve this, we first provide a systematization of the applications of wireless technologies upon which civil aviation relies. Based on these applications, we comprehensively analyze vulnerabilities, attacks, and countermeasures. We categorize the existing research on countermeasures into approaches that are applicable in the short term and research of secure new technologies deployable in the long term. Since not all of the required aviation knowledge is codified in academic publications, we additionally examine existing aviation standards and survey 242 international aviation experts. Besides their domain knowledge, we also analyze the awareness of members of the aviation community concerning the security of wireless systems and collect their expert opinions on the potential impact of concrete attack scenarios using these technologies.

Martin Strohmeier, Vincent Lenders, Ivan Martinovic

Automatic dependent surveillance-broadcast (ADS-B) is the communications protocol currently being rolled out as part of next generation air transportation systems. As the heart of modern air traffic control, it will play an essential role in the protection of two billion passengers per year, besides being crucial to many other interest groups in aviation. The inherent lack of security measures in the ADS-B protocol has long been a topic in both the aviation circles and in the academic community. Due to recently published proof-of-concept attacks, the topic is becoming ever more pressing, especially with the deadline for mandatory implementation in most airspaces fast approaching. This survey first summarizes the attacks and problems that have been reported in relation to ADS-B security. Thereafter, it surveys both the theoretical and practical efforts which have been previously conducted concerning these issues, including possible countermeasures. In addition, the survey seeks to go beyond the current state of the art and gives a detailed assessment of security measures which have been developed more generally for related wireless networks such as sensor networks and vehicular ad hoc networks, including a taxonomy of all considered approaches.