arXiv subjects

Konrad Rieck

Publications and source records attributed to Konrad Rieck.

At least 19 recordsLinked to original sources

arXiv2026

Secrets Everywhere: Auditing Memorization in Mobility Prediction Models

Human mobility prediction models, which forecast the next location in a user's trajectory, are increasingly deployed in urban analytics, navigation, and personalized services. Yet, little is known about their potential to memorize and expose sensitive user trajectories from training data. While memorization has been extensively studied in language models, mobility prediction poses unique challenges: training sequences encode human behavior at various spatial and temporal scales, creating privacy risks at different granularities. In this paper, we conduct the first systematic audit of memorization in mobility prediction models. While prior work has shown that privacy leaks can arise from such models, we systematically assess and quantify memorization risks at scale. We identify key challenges, including the lack of a randomness space, the multi-scale structure of trajectories, and user-specific behavioral diversity. To address these challenges, we introduce a framework to quantify mobility memorization at different levels of granularity: individual locations, anchor pairs, and subtrajectory segments. We also develop user-grounded reference sets to assess how likely a model is to prefer training data over realistic alternatives. Our evaluation across multiple models and datasets reveals pervasive memorization patterns that correlate with user regularity and increase the risk of data extraction at inference time. Our findings call for mandatory privacy auditing in mobility prediction models.

cs.LG↗

arXiv2026

False Prophets: On the Security of World Models in Agentic Systems

Large language models now power autonomous agents capable of complex, multi-step tasks in different environments. Accurate and reliable execution of these tasks requires the agent to predict the results of its actions. Recent research proposes to enhance predictive capabilities via specially trained environment simulators-world models. While world models can improve performance, they can also mislead agents into executing harmful actions, creating significant security and privacy risks. In this paper, we raise security concerns regarding the usage of world models in agentic systems. We discover a range of world model specific vulnerabilities, which can be exploited in terminal-based agents to execute malicious code or extract sensitive data. To facilitate future development, we introduce a security benchmark dataset designed for text-based world models. We argue that some risks are intrinsic to approximate world modeling, and show that attackers can induce mispredictions in agentic pipelines with up to 95% success rate, possibly resulting in unintended command execution, denial of service, drainage of wallet and private information extraction. Finally, we provide practical recommendations for practitioners to mitigate the discovered harms and harden agentic systems.

cs.CR↗

arXiv2026

Practical Type Inference: High-Throughput Recovery of Real-World Structures and Function Signatures

The recovery of types from stripped binaries is a key to exact decompilation, yet its practical realization suffers. For composite structures in particular, both layout and semantic fidelity are required to enable end-to-end reconstruction. Many existing approaches either synthesize layouts or infer names post-hoc, which weakens downstream usability. This is further aggravated by an excessive runtime overhead that is especially prohibitive in automated environments. We present XTRIDE, an improved n-gram-based approach that focuses on practicality: highly optimized throughput and actionable confidence scores allow for deployment in automated pipelines. When compared to the state of the art in struct recovery, our method achieves comparable performance while being between 70 and 2300 times faster. As our inference is grounded in real-world types, we achieve the highest ratio of fully-correct struct layouts. With an optimized training regimen, our model outperforms the current state of the art on the DIRT dataset by 5.09 percentage points, achieving 90.15% type inference accuracy overall. Furthermore, we show that n-gram-based type prediction generalizes to function signature recovery: conducting a case study on embedded firmware, we show that this efficient approach to function similarity can assist in typical reverse engineering tasks.

cs.CR↗

arXiv2026

Fingerprinting Inference Systems of Large Language Models

The behavior of LLMs does not depend solely on the model itself. Components of the inference system, such as the inference engine, attention backend, and hardware platform, subtly influence how inputs are processed. These components differ in their implementations and thereby induce small numerical deviations across systems when running the same model. While prior work has established the theoretical existence of such deviations, their security implications have remained unexplored. In this paper, we show that these deviations are characteristic of specific components and propagate to observable textual outputs, exposing the inference system to any party that can query the model. Building on this observation, we introduce a fingerprinting method that analyzes the prompt-response behavior of LLMs to identify components of the inference system. Our empirical evaluation demonstrates that the inference engine, attention backend, and underlying hardware platform can be identified reliably, even when the LLM is operated at non-zero temperature. We show that preventing fingerprinting is fundamentally hard, as it would require eliminating numerical differences between hardware and software stacks. We therefore propose partial mitigations and discuss their impact.

cs.CR↗

arXiv2026

What Are We Measuring in NLG? A Meta-Analysis of Evaluation Trends 2020-2025

As Natural Language Generation (NLG) dominates modern NLP, scalable evaluation remains a critical bottleneck. Consequently, LLM-as-a-judge (LaaJ) adoption has accelerated rapidly, appearing in more papers than human evaluation in 2025. This pivotal shift motivates a critical analysis of current evaluation practices. Overcoming the limits of rigid keyword filtering and manual review, we employ a multi-LLM information extraction pipeline to gather structured metadata from 14,171 papers across four major NLP conferences (2020-2025). Analyzing 3,334 filtered NLG papers, we identify three systemic challenges. (1) Metric inertia: despite the shift toward open-ended generation, legacy lexical metrics (BLEU, ROUGE) persist as primary indicators, typically used alongside rather than replaced by semantic alternatives. (2) Metric-criteria mapping problem: our paper-level co-occurrence data reveals that general-purpose automatic metrics are applied as broad proxies for quality, without specifying which dimension of text generation they are intended to evaluate. (3) Validation gap: LaaJ has grown rapidly without commensurate human validation (fewer than 8% of papers). Crucially, while LaaJ correlates with aggregate quality, alignment collapses on fine-grained criteria like fluency. To address these gaps, we distill our findings into a minimal Evaluation Checklist to guide metric selection, construct validity, and LaaJ deployment.

cs.CL↗

arXiv2026

Measuring Security Without Fooling Ourselves: Why Benchmarking Agents Is Hard

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.

cs.CR↗

arXiv2026

When a Zero-Shooter Cheats: Improving Age Estimation via Activation Steering

Different age-related regulations have been proposed to protect minors from harmful content and interactions online. Automated age estimation is central to enforcing such regulations, and vision-language models (VLMs) achieve state-of-the-art performance on this task. However, we find that the zero-shot nature of VLM-based age estimation produces an unexpected side effect we call the identity shortcut: Instead of estimating age from visual features, VLMs tend to identify the depicted person and infer their age from memorized knowledge. This phenomenon leads to substantially incorrect predictions when non-celebrities are misidentified as celebrities. It also produces deceptively high robustness to noise and adversarial perturbations on celebrity images, which dominate popular benchmarks. To mitigate this, we propose an activation steering method that suppresses the shortcut by intervening on the hidden states of the VLM. This method improves age estimation accuracy for both memorized and unseen identities, reducing mean absolute error by up to 25% across popular benchmarks.

cs.LG↗

arXiv2026

Toward Securing AI Agents Like Operating Systems

Autonomous agents based on large language models (LLMs) are rapidly emerging as a general-purpose technology, with recent systems such as OpenClaw extending their capabilities through broad tool use, third-party skills, and deeper integration into user environments. At the same time, these agentic systems introduce substantial security risks by combining unconstrained capabilities with access to sensitive user data. In this work, we investigate the security of LLM-based agents through the lens of operating systems. We argue that both face strikingly similar challenges in isolating resources, separating privileges, and mediating communication. Guided by this perspective, we survey the current landscape of open-source agents, derive a unified agent architecture, and systematically analyze potential attack vectors. To validate this analysis, we conduct a case study evaluating four widely used OpenClaw-like agents. Even under modest attacker capabilities, we find that several protection mechanisms fail in practice and that secure operation requires detailed system knowledge and careful configuration. However, we also observe that while some agentic capabilities remain insecure by design, many vulnerabilities can be mitigated using well-established techniques from operating system security. We conclude with a set of recommendations for the secure design of agentic systems.

cs.CR↗

arXiv2026

Almost for Free: Crafting Adversarial Examples with Convolutional Image Filters

Adversarial examples in machine learning are typically generated using gradients, obtained either directly through access to the model or approximated via queries to it. In this paper, we propose a much simpler approach to craft adversarial examples, drawing inspiration from insights of explainable machine learning. In particular, we design \emph{adversarial image filters} that are based on classic edge detection algorithms but optimized to deceive learning models. The resulting untargeted attacks are transferable and require only a single pass over the input. Empirically, we find that 3x3 filters already enable success rates between 30% and 80% on different neural networks. Compared to related approaches using generative models for crafting adversarial examples, we reduce the number of parameters by five orders of magnitude, resulting in a very efficient attack. When investigating the parameters of the learned filters, we observe interesting properties such as a high transferability between models and structures common to classic image filters. Our results provide further insights into the vulnerability of neural networks and their fragility to malicious noise.

cs.LG↗

arXiv2026

Hardware-Triggered Backdoors

Machine learning models are routinely deployed on a wide range of computing hardware. Although such hardware is typically expected to produce identical results, differences in its design can lead to small numerical variations during inference. In this work, we show that these variations can be exploited to create backdoors in machine learning models. The core idea is to shape the model's decision function such that it yields different predictions for the same input when executed on different hardware. This effect is achieved by locally moving the decision boundary close to a target input and then refining numerical deviations to flip the prediction on selected hardware. We empirically demonstrate that these hardware-triggered backdoors can be created reliably across common GPU accelerators. Our findings reveal a novel attack vector affecting the use of third-party models, and we investigate different defenses to counter this threat.

cs.LG↗

arXiv2026

Manipulating Feature Visualizations with Gradient Slingshots

Feature Visualization (FV) is a widely used technique for interpreting concepts learned by Deep Neural Networks (DNNs), which synthesizes input patterns that maximally activate a given feature. Despite its popularity, the trustworthiness of FV explanations has received limited attention. We introduce Gradient Slingshots, a novel method that enables FV manipulation without modifying model architecture or significantly degrading performance. By shaping new trajectories in off-distribution regions of a feature's activation landscape, we coerce the optimization process to converge to a predefined visualization. We evaluate our approach on several DNN architectures, demonstrating its ability to replace faithful FVs with arbitrary targets. These results expose a critical vulnerability: auditors relying solely on FV may accept entirely fabricated explanations. To mitigate this risk, we propose a straightforward defense and quantitatively demonstrate its effectiveness.

cs.LG↗

arXiv2025

LLM-based Vulnerability Discovery through the Lens of Code Metrics

Large language models (LLMs) excel in many tasks of software engineering, yet progress in leveraging them for vulnerability discovery has stalled in recent years. To understand this phenomenon, we investigate LLMs through the lens of classic code metrics. Surprisingly, we find that a classifier trained solely on these metrics performs on par with state-of-the-art LLMs for vulnerability discovery. A root-cause analysis reveals a strong correlation and a causal effect between LLMs and code metrics: When the value of a metric is changed, LLM predictions tend to shift by a corresponding magnitude. This dependency suggests that LLMs operate at a similarly shallow level as code metrics, limiting their ability to grasp complex patterns and fully realize their potential in vulnerability discovery. Based on these findings, we derive recommendations on how research should more effectively address this challenge.

cs.CR↗

arXiv2025

Adversarial Observations in Weather Forecasting

AI-based systems, such as Google's GenCast, have recently redefined the state of the art in weather forecasting, offering more accurate and timely predictions of both everyday weather and extreme events. While these systems are on the verge of replacing traditional meteorological methods, they also introduce new vulnerabilities into the forecasting process. In this paper, we investigate this threat and present a novel attack on autoregressive diffusion models, such as those used in GenCast, capable of manipulating weather forecasts and fabricating extreme events, including hurricanes, heat waves, and intense rainfall. The attack introduces subtle perturbations into weather observations that are statistically indistinguishable from natural noise and change less than 0.1% of the measurements - comparable to tampering with data from a single meteorological satellite. As modern forecasting integrates data from nearly a hundred satellites and many other sources operated by different countries, our findings highlight a critical security risk with the potential to cause large-scale disruptions and undermine public trust in weather prediction.

cs.CR↗

arXiv2025

Dancer in the Dark: Synthesizing and Evaluating Polyglots for Blind Cross-Site Scripting

Cross-Site Scripting (XSS) is a prevalent and well known security problem in web applications. Numerous methods to automatically analyze and detect these vulnerabilities exist. However, all of these methods require that either code or feedback from the application is available to guide the detection process. In larger web applications, inputs can propagate from a frontend to an internal backend that provides no feedback to the outside. None of the previous approaches are applicable in this scenario, known as blind XSS (BXSS). In this paper, we address this problem and present the first comprehensive study on BXSS. As no feedback channel exists, we verify the presence of vulnerabilities through blind code execution. For this purpose, we develop a method for synthesizing polyglots, small XSS payloads that execute in all common injection contexts. Seven of these polyglots are already sufficient to cover a state-of-the-art XSS testbed. In a validation on real-world client-side vulnerabilities, we show that their XSS detection rate is on par with existing taint tracking approaches. Based on these polyglots, we conduct a study of BXSS vulnerabilities on the Tranco Top 100,000 websites. We discover 20 vulnerabilities in 18 web-based backend systems. These findings demonstrate the efficacy of our detection approach and point at a largely unexplored attack surface in web security.

cs.CR↗

arXiv2025

Lazy Gatekeepers: A Large-Scale Study on SPF Configuration in the Wild

The Sender Policy Framework (SPF) is a basic mechanism for authorizing the use of domains in email. In combination with other mechanisms, it serves as a cornerstone for protecting users from forged senders. In this paper, we investigate the configuration of SPF across the Internet. To this end, we analyze SPF records from 12 million domains in the wild. Our analysis shows a growing adoption, with 56.5 % of the domains providing SPF records. However, we also uncover notable security issues: First, 2.9 % of the SPF records have errors, undefined content or ineffective rules, undermining the intended protection. Second, we observe a large number of very lax configurations. For example, 34.7 % of the domains allow emails to be sent from over 100 000 IP addresses. We explore the reasons for these loose policies and demonstrate that they facilitate email forgery. As a remedy, we derive recommendations for an adequate configuration and notify all operators of domains with misconfigured SPF records.

cs.CR↗

arXiv2025

On the Role of Pre-trained Embeddings in Binary Code Analysis

Deep learning has enabled remarkable progress in binary code analysis. In particular, pre-trained embeddings of assembly code have become a gold standard for solving analysis tasks, such as measuring code similarity or recognizing functions. These embeddings are capable of learning a vector representation from unlabeled code. In contrast to natural language processing, however, label information is not scarce for many tasks in binary code analysis. For example, labeled training data for function boundaries, optimization levels, and argument types can be easily derived from debug information provided by a compiler. Consequently, the main motivation of embeddings does not transfer directly to binary code analysis. In this paper, we explore the role of pre-trained embeddings from a critical perspective. To this end, we systematically evaluate recent embeddings for assembly code on five downstream tasks using a corpus of 1.2 million functions from the Debian distribution. We observe that several embeddings perform similarly when sufficient labeled data is available, and that differences reported in prior work are hardly noticeable. Surprisingly, we find that end-to-end learning without pre-training performs best on average, which calls into question the need for specialized embeddings. By varying the amount of labeled data, we eventually derive guidelines for when embeddings offer advantages and when end-to-end learning is preferable for binary code analysis.

cs.LG↗

arXiv2025

SoK: Where to Fuzz? Assessing Target Selection Methods in Directed Fuzzing

A common paradigm for improving fuzzing performance is to focus on selected regions of a program rather than its entirety. While previous work has largely explored how these locations can be reached, their selection, that is, the where, has received little attention so far. A common paradigm for improving fuzzing performance is to focus on selected regions of a program rather than its entirety. While previous work has largely explored how these locations can be reached, their selection, that is, the where, has received little attention so far. In this paper, we fill this gap and present the first comprehensive analysis of target selection methods for fuzzing. To this end, we examine papers from leading security and software engineering conferences, identifying prevalent methods for choosing targets. By modeling these methods as general scoring functions, we are able to compare and measure their efficacy on a corpus of more than 1,600 crashes from the OSS-Fuzz project. Our analysis provides new insights for target selection in practice: First, we find that simple software metrics significantly outperform other methods, including common heuristics used in directed fuzzing, such as recently modified code or locations with sanitizer instrumentation. Next to this, we identify language models as a promising choice for target selection. In summary, our work offers a new perspective on directed fuzzing, emphasizing the role of target selection as an orthogonal dimension to improve performance.

cs.SE↗

arXiv2025

Evil from Within: Machine Learning Backdoors through Hardware Trojans

Backdoors pose a serious threat to machine learning, as they can compromise the integrity of security-critical systems, such as self-driving cars. While different defenses have been proposed to address this threat, they all rely on the assumption that the hardware on which the learning models are executed during inference is trusted. In this paper, we challenge this assumption and introduce a backdoor attack that completely resides within a common hardware accelerator for machine learning. Outside of the accelerator, neither the learning model nor the software is manipulated, so that current defenses fail. To make this attack practical, we overcome two challenges: First, as memory on a hardware accelerator is severely limited, we introduce the concept of a minimal backdoor that deviates as little as possible from the original model and is activated by replacing a few model parameters only. Second, we develop a configurable hardware trojan that can be provisioned with the backdoor and performs a replacement only when the specific target model is processed. We demonstrate the practical feasibility of our attack by implanting our hardware trojan into the Xilinx Vitis AI DPU, a commercial machine-learning accelerator. We configure the trojan with a minimal backdoor for a traffic-sign recognition system. The backdoor replaces only 30 (0.069%) model parameters, yet it reliably manipulates the recognition once the input contains a backdoor trigger. Our attack expands the hardware circuit of the accelerator by 0.24% and induces no run-time overhead, rendering a detection hardly possible. Given the complex and highly distributed manufacturing process of current hardware, our work points to a new threat in machine learning that is inaccessible to current security mechanisms and calls for hardware to be manufactured only in fully trusted environments.

cs.CR↗