Papers

Blockchains have evolved from simple distributed ledgers into programmable platforms that process complex application logic and carry significant financial value. All modern Web3 systems share a common goal: providing secure, decentralized, and trustworthy execution in an increasingly interconnected environment. However, this evolution has shifted the attack surface from isolated infrastructure disruptions to programmable economic abuse and cross-domain exploits. In this article, we focus on the research of blockchain attacks and defenses. In particular, we categorize the threat landscape and corresponding mitigation strategies according to both a four-tier layered architecture (network, cryptographic, consensus, and application) and cross-domain trust boundaries. We seek to answer these important questions: How has the research in blockchain security evolved over the past decade, especially with the rise of decentralized finance (DeFi) and cross-chain interoperability? How do local security assumptions fail when protocols are composed, and what are the driving needs for Web3 security research in the future?

For decades, the National Vulnerability Database (NVD), the "Cathedral", has been the reference source for vulnerability information for downstream research and industry tasks, e.g., software update prioritization. An emerging "Bazaar" of diverse CVE Numbering Authorities (CNAs) has created many alternative and sometimes diverging sources. We conduct a systematic analysis of divergence in Common Vulnerability Scoring System (CVSS) metrics covering the NVD and the public CNAs. We also check for self-divergence: two identical textual descriptions of CVEs with identical CWEs are rated differently by the same CNA. The odds of diverging are widespread, not uniform and sometimes unexpected. The assessment of Attack Complexity, User Interaction, and Impact are the major metrics where divergence happens. To understand the root causes, we perform a qualitative study by reaching out to the NVD and other CNAs (both open sources and proprietary products). We also discussed the findings at the CVSS Special Interest Group of FIRST, the community responsible for maintaining and evolving the CVSS standard. The key insights are that while something might be due to human errors, in some cases diverging is actually the right thing to do and might require changes in the way CVEs are generated industry-wide, in other cases explaining divergence requires access to additional FAQs. The good news is that the situation is improving since 2025, the bad news is that if one downloads the whole NVD (or another CNA dataset) from several years and uses it for predictions, the models trained on one source do not reliably generalize to a different source (accuracy can drop by 40%). We discuss the implications for practice and research.

Arm Permission Overlay Extension (POE) is an intra-process isolation mechanism based on memory protection keys. This mechanism partitions virtual memory into regions whose access permissions can be reconfigured without invoking higher exception levels. Because POE does not enforce privilege separation, POE-based protection domains are applicable only as a software fault-isolation technique. The POE2 is an isolation mechanism that substantially extends POE1. Dealing with memory protection keys, it also introduces new registers and tables that regulate the operations permitted for code executing within protection domains associated with different keys. A permission is determined by the spatial index of the executing code (i.e., the protection key associated with code), the spatial index of the accessed memory, the operation type, and a new attribute of the thread context: its temporal index. POE2 provides stronger security guarantees than POE1, but also introduces considerable architectural complexity. Effective permissions arise from the interaction of multiple hardware components, making direct programming non-trivial and prone to subtle errors. In the paper, we present a detailed analysis of POE2 and introduce a universal programming model with a strong security model for POE2-based systems. The model abstracts the complexity of spatial and temporal indices while enabling typical patterns of partitioned software constructed using intra-process isolation.

This paper presents a novel impersonation attack framework that aims to fool RF Fingerprinting (RFFP) identification systems by synthesizing signals that replicate the hardware-specific impairments of a target device. Our framework leverages unsupervised learning to enable accurate impairment estimation, combined with signal processing-based generation to synthesize high-fidelity adversarial signals. Unlike prior works that assume full access to the legitimate (victim) RFFP classifier, we consider a more realistic attack strategy where the adversary performs the attack from a completely different transceiver hardware. We further evaluate our proposed attack under realistic and challenging deployment settings, including over-the-air transmission in both Line-of-Sight (LoS) and Non-Line-of-Sight (NLoS) scenarios. Extensive experiments conducted on a Bluetooth Low Energy (BLE) device testbed demonstrate that our attacks remain highly effective even under severe access constraints, significantly outperforming existing baselines in terms of targeted attack success rates by over 80%. We additionally analyze the effects of cross-domain generalization, signal representation mismatch, and classifier diversity, highlighting the robustness and

AI agents issue tool calls on the basis of text they cannot verify, so any party who controls part of the context can forge the appearance of authority. I evaluate 15 contemporary language models against eight attack scenarios derived from a published corpus of real agent incidents and find that refusal varies from 100% down to 38% across fully evaluated models; the most expensive model refused only half of the attacks despite a twentyfold price spread. I present aiAuthZ, an authorization gateway that moves the safety decision off the agent's host. Before a tool call executes, the gateway verifies caller identity with a per-message HMAC-SHA256 signature bound to a single-use nonce and a timestamp window, and it evaluates a role-based and argument-level policy that the agent can neither read nor modify. Every decision joins a SHA-256 hash-chained audit log, and each accepted message yields an HMAC-authenticated QR receipt that achieves 94% mean verification across eight transmission channels, with zero forgeries accepted in 25 wrong-key trials. With the gateway in place, residual attack success falls to 0% for all 15 models at no more than 0.03 ms of added decision latency. On the AgentDojo banking suite, aiAuthZ blocks all seven attacker-directed tool calls the evaluated agents emit, at the cost of one legitimate first-time payment, while a spotlighting baseline allows two injections to succeed. Across nine in-scope case studies from the same incident corpus, aiAuthZ blocks nine of nine, against four of nine for a policy baseline without identity binding. The gateway does not prevent a model from being deceived; it prevents a deceived model from acting beyond the verified user's authority on every call routed through it. The implementation and all experiments are released at https://github.com/Sports-Vision-Inc/aiAuthZ.

Model-specific adversarial attacks have been extensively studied. We study a different failure mode: naturally occurring statistical signals in vision data that can behave as backdoor-like triggers without being maliciously inserted. We call these signals statistical adversaries. We analyse ImageNet to find patterns that are strongly linked to certain labels. We then use statistical controls to remove random correlations from our candidate signals. Finally, we demonstrate that these signals directly and predictably alter model predictions. These statistical adversaries are more targeted than generic corruptions and transfer across different model architectures. This suggests that some vulnerabilities are driven by dataset structure and distribution rather than a single model's idiosyncrasies. We conclude that ordinary datasets can contain exploitable adversarial surfaces even in the absence of poisoning, and suggest that dataset audits should treat spurious structure not only as a source of bias or interpretability failure, but also as a latent attack surface for vision models.

Detection models running in adversarial environments face a malicious distribution that drifts rapidly while the benign distribution stays comparatively stable, so teams retrain and redeploy constantly to stay ahead of new threats. Retraining tends to change the output prediction scores, which breaks downstream users of the model. For these security-oriented models we need consistent false-positive rate (FPR) across all output values, whereas standard probability-calibration methods target class probability rather than an FPR contract. We introduce a method built on top of existing calibration primitives that targets the whole FPR curve, giving scores a consistent FPR meaning across deployments. On one held-out split, the observed relative FPR error was at most 2.3% from 10% down to 0.1% FPR and 7.2% at 0.01% FPR. The shipped artifact remains under 200 KB in measurements across calibration sets from 1K to 10M benign samples.

Generative AI (GenAI) systems store and process client data in three distinct ways: in the model's parameters through training and memorisation, in the context window during a live session, and in knowledge databases for retrieval-augmented generation (RAG). Each mode creates different and often counter-intuitive risks to confidentiality and legal professional privilege, and each calls for specific governance responses. Drawing on the first English and American decisions to address privilege and generative AI, UK and Munir v Secretary of State for the Home Department and United States v Heppner, on the orthodox privilege authorities against which those decisions must be read, and on recent computer science research, we explain the three modes of data storage and processing in terms accessible to practitioners and analyse the legal consequences of each. We then situate the analysis within the regulatory framework governing solicitors in England and Wales and within the ordinary principles of professional negligence, arguing that the standard of effective information governance (and with it the benchmark against which negligence and misconduct will be measured) is changing. Although we write primarily for SRA-regulated practitioners, our data-governance analysis is framed to extend to any jurisdiction in which the protection of privilege or professional secrecy depends on demonstrable confidentiality. The ultimate aim of this article is to help legal services professionals understand salient data leakage risks in GenAI systems and thereby facilitate a more responsible deployment of GenAI on client data and other sensitive material.

Algorithmic Complexity Vulnerabilities (ACVs) arise when adversarial inputs trigger worst-case execution behavior, causing severe performance degradation or Denial-of-Service conditions. A key but underexplored source is shadow complexity: non-trivial computational costs hidden inside seemingly benign standard library APIs. Because these costs are invisible at call sites, attackers can exploit them to induce unexpected superlinear runtime behavior. Existing ACV detectors often rely on fuzzing, symbolic execution, or hybrid analysis, but they are usually language-specific, require substantial manual effort to construct harnesses, and depend on heavy runtime instrumentation. We present ShadowProbe, a scalable and language-extensible framework for discovering ACVs through lightweight static analysis, automated reconstruction of execution contexts, and Large Language Model (LLM) assisted test generation. ShadowProbe uses a structured multi-stage pipeline: it statically screens for candidate functions guided by shadow-complexity signals, reconstructs minimal executable contexts from project-level symbols, and synthesizes size-controlled inputs to probe worst-case behavior. It then validates candidates using execution-time measurements and robust statistical growth inference, separating true algorithmic blowups from runtime noise such as garbage collection and JIT compilation effects. We evaluate ShadowProbe on the WISE benchmark, where it consistently improves analysis efficiency over existing approaches. We further apply it to large-scale systems including CPython, the JDK, Zig, Rustc, and vLLM, uncovering many previously unknown ACVs, many of which have been confirmed and partially remediated by maintainers. These results show that ShadowProbe can identify hidden algorithmic risks across diverse real-world codebases.

As AI agents are incorporated into life science workflows, the capabilities that speed discovery might also enable misuse. We present BioSecBench-Refusal, a benchmark for risk identification and refusal behavior for biological research tasks. The benchmark pairs 61 Routine tasks, legitimate analyses adapted from the published literature, with 46 Red-Team tasks, fictional scenarios that resemble real research but conceal a biosecurity hazard. Across 16 model-harness configurations, refusal rates ranged from 7 percent to 74 percent on Routine tasks and 1 percent to 62 percent on Red-Team tasks, with many configurations refusing legitimate Routine work at comparable or higher rates than concealed hazards. Refusals were most often triggered by provider API filters applied prior to agentic reasoning. However, models given room to reason showed the potential to identify more real threats. We release BioSecBench-Refusal as a tool for model developers to calibrate capability and caution for agentic biotech research and development.

Watermarking methods embed imperceptible and verifiable signals into text generated by large language models (LLMs). Existing approaches include zero-bit schemes for distinguishing synthetic text from human writing and multi-bit schemes for embedding metadata. However, current multi-bit watermarking methods do not allow selective disclosure: verifying any part of the watermark requires revealing the entire embedded message. This lack of control leads to unnecessary information exposure and raises privacy concerns. We propose Hierarchical Vocabulary Routing (HeRo), a watermarking framework that enables selective disclosure of embedded metadata. The method recursively partitions the vocabulary and distributes watermark information across hierarchical layers, so that different verifiers can decode only the portions of the payload corresponding to their access level. We show that the proposed scheme preserves the unbiasedness of the underlying sampling process and thus maintains text quality. Experiments demonstrate that our framework supports fine-grained access control while achieving high detection accuracy and low latency. Code is available at https://github.com/xuyangc03/hero-watermark.

As LLM agents evolve from single-user assistants into shared organizational infrastructure, new privacy risks emerge: inappropriate information may not only be exposed through outputs for external recipients, but also internally across users through inter-agent messages, shared memory and agents. These data spillage risks are not captured by existing privacy benchmarks grounded in contextual integrity (CI) as they focus primarily on either single-user settings or interactions between independently owned agents. We introducePiSAs (Privacy in Shared Agentic systems), a benchmark for assessing unintentional leaks with dual CI annotations: whether an information is appropriate for the task, and which users may legitimately access it. This enables direct measurement of cross-user spillage across agentic system components and interfaces, such as outputs, inter-agent communication, and memory. PiSAsis system-agnostic and supports evaluation across different agent topologies and memory regimes. We find that, although system design improves CI compliance, results are bottlenecked by incorrect LLM judgment calls: even state-of-the-art models fail to reliably filter inappropriate content or restrict transmission to authorized users. Our findings underscore the need for privacy-preserving strategies, beyond those studied in this work.

Parameter-efficient fine-tuning still leaves a broad space of behavior-changing updates reachable, so a poisoned objective can be represented and optimized. We study an alternative: adaptation constrained to the subspace estimated from a trusted pool of existing task adapters. On flan-t5-large with 196 public LoRA adapters, we show that (1) the functionally relevant content of an adapter lies in a low-dimensional shared subspace, 30 to 38 percent of its weight norm being redundant under the evaluated task distributions; (2) gradient adaptation restricted to 128 coordinates on this subspace matches full LoRA fine-tuning on clean classification data, while under targeted label inversion LoRA collapses to 3-26 percent exact match and the constrained learner keeps 62-96 percent on the tasks the pool covers; (3) the constrained learner cannot fit corrupted data, its adaptation loss separating clean from garbage by two orders of magnitude (120x), an out-of-distribution signal without an extra detector; and (4) against an adaptive backdoor attacker who optimizes within the subspace, the attack is blocked (8 percent success versus 100 for LoRA) on the task where its target behavior is unlike anything in the pool, and only partially blocked (85 percent) when the target coincides with a common pool behavior. On these two tasks the outcome is consistent with how close the target is to the pool's directions, which suggests but does not establish a pool-relative boundary. The mechanism trades peak plasticity for these properties: on tasks the pool covers poorly, unconstrained fine-tuning wins, and the protection assumes the pool itself is trusted. Code and data are public.

Defenses that provide security guarantees against prompt injection attacks rely on strict isolation between trusted instructions and untrusted data. In text-based environments such as tool-use APIs, this separation arises naturally: agents can reason from interface definitions without ever processing untrusted content. Extending these guarantees to web agents faces a fundamental challenge: to perceive and interact with their environment, web agents must first observe the rendered page, which intermingles trusted content with untrusted content. This structural entanglement removes the trust boundary on which security guarantees depend, undermining provable defenses for web agents. In this paper, we present Untrusted Content Masking (UCM), a simple and effective approach that restores this boundary in web environments. We leverage a key structural insight: a webpage's Document Object Model (DOM) encodes sufficient information to distinguish trusted from untrusted regions without reading their content. Our framework exploits this by redacting untrusted regions before they reach the agent and routing interaction through a sandboxed interface with strict privilege separation, thereby enabling agents to observe and interact with their environment while remaining isolated from adversarial content. The code is publicly available.

Neural network verification and data privacy are inherently in tension: verification demands full access to model parameters and input data, yet both are increasingly restricted by privacy regulations and intellectual property constraints. This tension has left robustness verification impractical in privacy-sensitive domains. In this work, we address this gap with SecureCROWN, the first framework for privacy-preserving neural network robustness verification. Built upon secure two-party computation (2PC), our framework enables a model owner and a data owner to jointly compute certified robustness bounds – revealing only the final result while provably protecting both parties' private data under the semi-honest security model. A key challenge is securely computing the conditional operations in Linear Bound Propagation, where the data-dependent branching is incompatible with standard secure computation protocols. We eliminate branching by formulating conditional logic as continuous arithmetic operations. Additionally, we introduce a Newton–Raphson refinement method to improve numerical stability. Extensive analysis and experiments show that SecureCROWN strictly matches plaintext verification results, while completing in 0.1–200s across varied model sizes and communication settings (LAN/WAN), demonstrating the feasibility of privacy-preserving neural network verification.

Persistent personal agents combine long-term memory with access to users' external environments, enabling personalized foreground assistance and proactive background execution. This integration also creates a new path to compromise: untrusted external content can be silently written into persistent memory and later reused as trusted state. We study this threat as stealth memory injection, in which a remote black-box adversary delivers a single email payload that must induce the agent to write poisoned memory, stay hidden in the agent's response to the user, and affect future behavior. We introduce WhisperBench, a 108-case benchmark spanning five risk categories and both fact and preference poisoning. Built on a real IMAP/SMTP workflow and an authentic email agent skill, it enables full-cycle evaluation of stealth memory injection attacks. To enable this black-box attack under single-email delivery and without runtime feedback, we propose MemGhost, a one-shot payload generation framework. MemGhost uses an environment proxy to emulate persistent-agent execution and an objective proxy to convert memory adoption and conversational stealth into dense rubric-based rewards, then trains the attacker policy with supervised fine-tuning and reinforcement learning. Across 56 held-out test cases, MemGhost achieves 87.5% end-to-end success on OpenClaw with GPT-5.4 and 71.4% on Claude Code SDK with Sonnet 4.6. It also transfers across personal-agent architectures (NanoClaw and Hermes Agent) and memory backends (filesystem and vector-based Mem0), and remains effective against input-level, model-level, and system-level defenses. These results suggest that persistent memory can turn ordinary external processing into a practical pathway for long-term agent compromise.

We present a new algebraic modeling of the Supersingular Isogeny Problem as a system of multivariate polynomial equations, in the case where the elliptic curves are connected by an isogeny whose degree is a power of $2$ or $3$. This modeling relies on Renes formulas for elliptic curves in Montgomery form (degree $2$) or triangular form (degree $3$). We investigate several algebraic properties of these systems: we prove that they are zero-dimensional, compute the dimension of their highest degree part, and show that they are not in generic coordinates. Experimental results show that solving these systems via Gröbner basis techniques is significantly faster than solving the algebraic modeling with modular polynomials.

AI agents act on behalf of user prompts, consuming external data and taking actions based on the agent context. Prior research on AI agent security has primarily focused on indirect prompt injection (IPI). Its most well-studied category is instruction injection, where attacker-controlled untrusted data is interpreted as an instruction. In response, many mitigations have been proposed to prevent instruction injection attacks. In this paper, we introduce a new category of IPI, agent data injection attacks (ADI). ADI injects malicious data disguised as trusted data, such as security-critical metadata (e.g., resource identifiers or data origins) or agent context data (e.g., tool call and response formats). As a result, agents unknowingly execute unintended actions based on attacker-controlled data. ADI has similar attack impacts as instruction injection attacks, because it causes agents to misbehave and execute unintended actions. Despite the similar impact, ADI remains underexplored and easily bypasses existing IPI defenses. We found several critical vulnerabilities in real-world agents that allow an attacker to launch various attacks: arbitrary click attacks on web agents (Claude in Chrome, Antigravity, and Nanobrowser), and remote code execution and supply-chain attacks on coding agents (Claude Code, Codex, and Gemini CLI). We evaluate ADI vulnerabilities across off-the-shelf models and AI agents, and find that ADI is effective in both standalone LLMs and AI agent settings. ADI exposes a critical gap in agent security, signifying that current AI agents do not employ a fundamental security principle: current agents do not isolate trusted data from untrusted data.

To efficiently exploit a valuable data source (e.g., facial or medical images), it is frequently harnessed to fulfill multiple learning objectives (e.g., facial recognition, age estimation, and race classification). Each trained model is then deployed as an independent API service for corresponding inference. However, the privacy risk introduced by this one-dataset-multiple-model (ODMM) paradigm is completely overlooked by the community. For the first time, this work reveals that the ODMM setting substantially amplifies privacy leakage. We establish a theoretical framework that proves that privacy leakage accumulates as more ODMM models are exposed, a phenomenon we term ODMM privacy composition. Guided by this theoretical foundation, we propose PRIME (Privacy Amplification RIsk from One-Dataset-Multiple-Model Exposure) to systematically assess this risk and quantify the resulting leakage using membership inference attacks (MIAs). Under black-box access to ODMM models, we design an aggregation mechanism that collectively captures carefully identified privacy signals leaked by individual ODMM models, and construct an attack meta-classifier over the aggregated meta-information to infer the membership status of a given sample jointly. Our results provide strong evidence that dataset reuse across ODMM models strikingly jeopardizes privacy, which is consistently evident across five privacy-sensitive image and textual benchmark datasets and diverse model architectures (from ResNet and ViT to Qwen3-1.7B), spanning three domains: facial analysis, medical imaging, and textual attribution analysis. While mitigations such as differential privacy can reduce the effectiveness of PRIME with trade-offs, our attack still consistently outperforms single-task MIAs.

In physical layer authentication, verification of a user's identity is based on the characteristics of the transmission channel through which signals are delivered to the authenticator (Bob). In this paper, we assume that the signals received by Bob pass through a \ac{RIS} (controlled by Bob) and that the legitimate transmitter (Alice) is equipped with one antenna. Conversely, the attacker (Trudy) has multiple antennas and uses precoding to deceive Bob's verification. Assuming that Trudy knows all the channel matrices, we first derive her optimal attack strategy. Then, we analyse the conditions under which the channel estimated by Bob is indistinguishable when either Alice or Trudy is transmitting. When Trudy has a single antenna, we show that the indistinguishability condition cannot be met when the channels to the RIS are the result of propagation over multiple paths. For single-path line-of-sight (LOS) conditions, instead, Trudy can impersonate Alice although transmitting from a different position. We verify these results numerically and assess the security of the considered scenario, even when the indistinguishability conditions cannot be met.

Persistent memory has enabled large language model (LLM) agents to store factual knowledge, prior decisions, reasoning histories, tool usage information, and context. While this has improved the agent's functionality and continuity across tasks, it has also introduced a new attack surface: the agent's own reasoning history. In this paper, we introduce the Forged Amplifying Rationale Memory Attack (FARMA), which poisons an agent's remembered reasoning rather than its factual knowledge. It inserts forged reasoning traces using evasive language that bypasses keyword-based defenses, then amplifies them through self-referential reinforcement that defeats consensus-based defenses. To address FARMA, we introduce SENTINEL, a layered defense pipeline to detect forged reasoning entries. Its central component is the Reasoning Guard that structurally analyzes candidate entries for forgery using five weighted signals. We evaluate FARMA and SENTINEL across multiple agents and different LLM models with 50 trials and show that FARMA achieves an attack success rate of up to 100% under baseline conditions and is capable of defeating defense mechanisms like keyword filter and A-MemGuard. Our evaluation also shows that SENTINEL reduces FARMA's attack success rate to as low as 0% with no false positives observed across 326 benign agent traces. Our work demonstrates the need to protect not only an agent's retrieved content but also the integrity of its reasoning history.

This paper develops a representation-theoretic perspective on cryptographic protocols. The focus is not solely on the computability of the abstract value as an extensional property, but on the algorithmic structure of its presentation in a representation system: for operational use in protocols, algorithmic accessibility of the value does not suffice; its fixed presentation is also decisive. We distinguish three representation-theoretic notions – algorithmically approximable (A_app, the computable real numbers), finitely exactly describable in a system (A_fin(S)), and canonical normalizability of a system – and show that there is no computable extensional canonicalizer that uniformly transforms arbitrary approximation programs of computable real numbers into unique finite value encodings. As the operational rational core presentation we use the rational system with its canonical encoding specification Sigma_Q (fixed rules for valid fraction descriptions, canonical codes, and normalization); the associated value set is A_ex = Q. The notion of a canonically serializable object class transfers this core idea to practical protocol objects (files as byte sequences, hash values, transaction IDs, and normatively serialized payloads). We illustrate the consequences for interoperability, well-definedness, and verification with fully worked toy examples from symmetric and asymmetric encryption and hashing, and with a real-world example, the snaproot hash-anchoring protocol for blockchain-based file integrity verification. The paper thereby shows that the mathematical determinacy of a value and its operational uniqueness as a protocol object are two different requirements. Once a normative representation specification has been fixed, byte-level correctness and well-definedness arguments can be carried out without further implementation-dependent serialization or rounding decisions.

Cyber Threat Intelligence (CTI) reports are predominantly unstructured, heterogeneous, and noisy, which limits their direct usability for automated analysis and reasoning. Cybersecurity Knowledge Graphs (CSKGs) provide a structured representation of adversarial entities, actions, and relations, but constructing such graphs from free-text CTI remains a challenge. Recent approaches rely on monolithic Large Language Models (LLMs) to perform end-to-end extraction and completion, leading to high cost, limited controllability, and unstable performance. This paper introduces TACTIC-KG, an agentic framework for CSKG construction that decomposes the task into modular, specialized LLM agents responsible for extraction, typing, verification, and curation. Using lightweight models (3B–8B), TACTIC-KG improves stability, recall, and graph consistency while reducing deployment cost. We implement and evaluate TACTIC-KG against recent state-of-the-art systems. Experiments on human-annotated CTI reports show that agent specialization consistently outperforms larger monolithic in-context-learning (ICL) baselines in extraction F1-score, typing accuracy, and structural graph similarity.

Systems that process medical data should be meticulously secured. Yet, network services in healthcare environments often fail to implement basic security measures. For example, previous studies showed that network segmentation flaws led to DICOM systems leaking millions of patient records. In addition to DICOM, healthcare facilities rely heavily on the HL7 and FHIR protocols to transmit data. For nine months, we operated a low-interaction honeypot for medical protocols. We found it was regularly scanned for DICOM but never for HL7 or FHIR, indicating that despite their widespread use and importance for patient data security, the security of these services remains underexplored. In this paper, we present the first large-scale study on HL7 and FHIR services and expand previous work on DICOM. Our large-scale Internet scans, covering the three major healthcare protocols across IPv4 and IPv6 address spaces, identify healthcare systems and uncover data leaks due to authentication flaws. Additionally, we scanned for deficiencies in TLS configurations of these services and known insecure healthcare software. In total, we found 2,841 healthcare services with authentication flaws. 94.4% of all exposed systems do not support transport encryption, and 1,373 systems have known software vulnerabilities, including those with potential for system takeover and CVSS scores up to 9.8. Overall, our study reveals an alarming state of cybersecurity in healthcare deployments, for which we discuss potential reasons and countermeasures. Finally, we report on the coordinated disclosure campaign we initiated to improve the security of patient data.

Look-ahead bias (using information from after a decision epoch to make the decision at that epoch) is the dominant way a backtest or a machine-learning evaluation flatters a system that will disappoint in deployment. The field manages it with construct-specific recipes and empirical detectors, which are sound only channel by channel and certify nothing by their silence. We show that look-ahead-freedom is a formal property in disguise: fixing an epoch, the demand that the future not influence the present is temporal non-interference over a time-indexed information lattice. From this identification we develop a pipeline calculus separating a datum's availability from its reference time, and settle the problem's boundary. Where availability may depend on data values, look-ahead-freedom is undecidable (indeed Pi-0-1-hard): leakage is recursively enumerable but freedom is not. On the value-independent fragment (covering windowing, resampling, joins, point-in-time and vintage reads, and agentic retrieval) we give a type-and-effect system that is sound and decidable in linear time. An artifact confirms the theory: the check scales linearly, an independent oracle witnesses no leak in any accepted pipeline, and the checker catches every planted leak that differential and tiling detectors miss.

We investigate the use of twin optical fingerprints derived from correlated physical unclonable functions (PUFs), as a hardware-based platform for cryptographic key generation and distribution. Each fingerprint is associated with a random, yet reproducible speckle pattern, generated when coherent light is scattered by a disordered optical structure. We consider a pair of correlated optical PUFs, and study the conditions under which two honest parties can establish a common secret key, despite fabrication-induced variability and environmental noise. An explicit information-theoretic key-agreement protocol is developed, incorporating secure sketches, error reconciliation, and privacy amplification. We quantify information leakage due to public helper data, and derive lower bounds on the length of the final secret key. The analysis identifies the noise regimes in which secure key agreement is feasible, and examines the performance of both practical and near-capacity reconciliation schemes. Finally, we discuss how twin optical PUFs could be integrated into quantum key distribution (QKD) networks, as a mechanism for establishing an initial pre-shared secret key between two honest users, without relying on computational assumptions or trusted third parties.

Fully homomorphic encryption (FHE) enables computation on encrypted data, but practical encrypted Transformer inference is bottlenecked by the sequential composition of many nonlinear blocks. We study whether Structured Newton Layer Parallelism (SNLP) can make this inter-layer composition more FHE-friendly: each Transformer block still requires polynomial approximations for operations such as softmax and RMSNorm, but SNLP reduces the layerwise sequential nonlinear depth from L stages to a small number of solver iterations plus linear structured corrections. Using a simulation framework based on Chebyshev polynomial approximations, we measure error accumulation under sequential versus SNLP inference across 8 models and 4 architecture families. On a 0.5B IDN-trained model, SNLP reduces symbolic bootstraps from 53 to 20 (2.65x) with only +1.2% perplexity degradation, while lowering error amplification (1.36x vs. 1.42x). Across all tested models, SNLP has lower amplification than sequential inference. Ablations show that softmax approximation dominates the error budget and CKKS arithmetic noise is negligible in our setting, suggesting that SNLP is complementary to block-level FHE-friendly operator design rather than a replacement for it.

Classical partially synchronous state machine replication, as in PBFT, tolerates f Byzantine replicas among n at least 3f+1 using three communication steps per request. Recent protocols such as Minimmit achieve two-message-delay decisions under stronger size assumptions, notably n at least 5f+1 when any silent replica must be counted as a potential equivocator. Hydrangea and Kudzu treat mixed Byzantine and crash faults, focusing on providing a fast-path under optimistic conditions while maintaining a fall-back commitment path similar to PBFT. In this paper, we also consider a mixed model, but focus on studying the fault tolerance of the 2-message-delay commit. For this, we prove a tight bound of n at least 5f+3c+1. Extending this result, we also show that there exists a more resilient commit path that allows an extra f_abc < n-3f-2c alive-but-corrupt faults at 4-message-delays. Core liveness is claimed in executions with at most f equivocators; if this regime is violated (e.g., AbC-induced forks), the protocol enters synchronous recovery, where only the resilient-path safety guarantee is preserved. As a result, for f=16, c=6, and n=99, we obtain a commit path that tolerates 22% of replicas failing for liveness, 16% equivocating for 1-RTT safety, and 54% equivocating for 2-RTT safety.

Cellular data billing is a core operational mechanism for mobile Internet service providers (ISPs), and a policy gap that excludes a specific protocol from usage accounting can lead to a practical security threat. Some cellular ISPs treat ICMP echo traffic as control traffic rather than user data and exclude it from billing. At the same time, Android allows ordinary applications to create ICMP echo sockets without root privileges because of an unsafe default configuration, and the combination of these two conditions forms a vulnerability that can bypass data billing. Existing billing-bypass attacks either require root privileges to create raw sockets and modify routing tables, or do not provide an end-to-end implementation that works in a non-rooted environment, which limits the threat to a small group of experts. This paper proposes Ghost Traffic, an end-to-end system that uses Android's VpnService to encapsulate all application traffic into ICMP echo payloads without root privileges and route it through an external proxy server. The proposed system targets both public IPv4 environments and IPv6-only LTE environments through two variants: IPv4 ICMP tunneling and IPv4-over-IPv6 ICMP tunneling. We evaluated its applicability in seven ISP environments in South Korea, Japan, and the United States, and observed end-to-end tunneling in six of them. We observed that billing bypass occurred in multiple environments and quantitatively showed this effect by measuring that Quality of Service (QoS) throttling was not applied even after the data cap was exhausted. Finally, we propose layered countermeasures across the device, platform, and network levels, performed responsible disclosure, and show that the operational practice of not billing ICMP traffic can lead to practical billing bypass.

Graph-structured data is increasingly generated and stored in decentralized environments, such as social platforms, mobile applications, and edge networks, where users maintain control over their local graph data. However, collecting and analyzing such decentralized graph data for downstream learning tasks raises significant privacy concerns, as nodes and their attributes often contain sensitive personal information. Local Differential Privacy (LDP) has emerged as a promising solution for privacy-preserving data collection without relying on trusted servers. Nevertheless, existing LDP-based graph learning methods typically assume uniform privacy requirements across users, ignoring the heterogeneous and personalized privacy preferences commonly observed in real-world systems. This uniform treatment leads to inflexible noise injection at the data collection stage, resulting in substantial distortion of graph data and degraded utility in subsequent analysis. To address this limitation, we propose PPGNN, a personalized differentially private framework for decentralized graph data. PPGNN enables user-specific privacy budgets during local perturbation while preserving analytical utility. To handle heterogeneous privacy levels and noise distortion, we design a two-stage solution consisting of a Personalized Perturbation Mechanism (PPM) and a weighted calibration strategy, FlexProp. Extensive experiments on six real-world graph datasets demonstrate that PPGNN effectively balances personalized privacy protection and data utility in decentralized graph learning scenarios.