Papers

Edge deployment of transformer-based models increasingly relies on ASIC accelerators due to their high performance and energy efficiency, achieved through optimized dataflows, specialized architectures, low-bitwidth computation, and efficient memory hierarchies. However, these advantages come with significant security vulnerabilities. ASIC-based DNN accelerators are susceptible to side-channel attacks (e.g., power, electromagnetic, and timing analysis) and fault injection attacks (e.g., voltage manipulation, clock glitches, and memory perturbations), which can lead to model extraction or compromised inference integrity. Furthermore, threats introduced during design and fabrication, such as hardware Trojans or untrusted third-party IPs, further expand the attack surface. To address these challenges, we explore a hybrid ASIC+eFPGA architecture that combines the efficiency of ASICs with the flexibility of reconfigurable logic. The integrated eFPGA enables security-oriented mechanisms such as adaptive runtime monitoring, side-channel mitigation and post-deployment patching. By leveraging these capabilities, the proposed approach enhances system resilience against both runtime and supply-chain attacks, while preserving the performance benefits of ASIC-based transformer inference.

We extend the CDPR's quantum attack from ideal lattices to module lattices over $2^k$-th cyclotomic rings. Using trace orthogonality of the power basis, we decompose a rank-$d$ module into mutually orthogonal rank-$1$ submodules, and apply CDPR's analysis to each independently and return the shortest candidate. The Hermite factor $\exp(\tilde{O}(\sqrt{n}))$ matches the ideal case, with a module reduction factor $\alpha_d=O(1)$ independent of the rank, under a balance hypothesis (proved for Gaussian distribution) automatic for MLWE-distributed bases. To enable a bounded-precision implementation, we replace coordinate-wise rounding with Chinese Remainder Theorem-scaled rounding at totally split primes, reducing the Gram-Schmidt rounding radius from $n/2$ to $\le 1$ at cost $O(d^2 r n \log n)$. Finally, we reformulate the CDPR's sign-selection step as a mixed-integer linear program and prove its optimum is no more than 1/2 for all $k$ ($\approx 0.4407$ for all tested $k\le 12$, conjecturally universal). This replaces the previous heuristic discrepancy $\Theta(\sqrt{nk})$. All results build on the class number condition $h_k^+=1$ established in Part I of this series.

Autonomous systems increasingly operate under partial observability where execution-relevant state is never fully accessible. Existing governance mechanisms – trusted execution environments, oracle-signed state proofs, cryptographic attestation – enforce the integrity of computation and state projections. We show this is structurally insufficient: an authenticated projection of state is necessary but never sufficient for execution validity. We introduce the Reconstructive Authority Model (RAM), which separates integrity from coverage. RAM defines a reconstruction gate that reasons over an explicit coverage envelope – comprising proven state, declared assumptions, and an acknowledged unobservable residual – and permits execution only when coverage is adequate for the action class. When coverage is insufficient, RAM narrows privileges dynamically or fails closed. Attestation proves trust in measurement; RAM proves adequacy of what is measured. We formalize RAM, prove necessity via two theorems (attestation insufficiency and RAM necessity) and three corollaries, and present a hybrid RAM+Attestation architecture with privilege-narrowing. Synthetic experiments (N=100,000, seed=42) show RAM achieves zero invalid execution rates at all coverage levels. Attestation-based systems exhibit IER=0.423 at low coverage and IER=0.233 even at full coverage, the latter arising from undefined-state handling failures undetectable by integrity checks alone. This reframes execution validity as a coverage reconstruction problem, distinct from and complementary to integrity guarantees provided by attestation.

Agent skills introduce a new and more severe form of indirect injection for LLM agents: unlike traditional indirect prompt injection, attackers can hide malicious instructions inside a dense, action-oriented skill that already functions as a legitimate instruction source. We study pre-execution skill-poison detection and show that successful skill poisoning induces a structured internal effect, attention hijacking, in which response-time attention shifts from trusted context to malicious skill spans and drives harmful behavior. Motivated by this mechanism, we propose RouteGuard, a frozen-backbone detector that combines response-conditioned attention and hidden-state alignment through reliability-gated late fusion. Across both real and synthetic open-source skill benchmarks, RouteGuard is consistently the strongest or most robust detector; on the critical Skill-Inject channel slice, it reaches 0.8834 F1 and recovers 90.51% of description attacks missed by lexical screening, showing that defending against skill poisoning requires internal-signal detection rather than text-only filtering

We identify and formalize a novel security risk: Context-Fragmented Violations (CFVs) - a class of policy breaches where individual agent actions appear locally safe and reasonable, yet collectively violate organizational policies because critical policy facts are siloed in different departments private contexts. Existing prompt-based alignment mechanisms and monolithic interceptors are poorly matched to violations that span contextual islands. We propose Distributed Sentinel, a distributed zero-trust enforcement architecture that introduces the Semantic Taint Token (STT) Protocol. Through lightweight sidecar proxies, our system propagates security state across organizational boundaries without exposing raw cross-domain data, enabling Counterfactual Graph Simulation for cross-domain policy verification. We construct PhantomEcosystem, a comprehensive benchmark comprising 9 categories of realistic cross-agent violation scenarios with adversarially balanced safe controls. On this benchmark, Distributed Sentinel achieves F1 = 0.95 with 106ms end-to-end latency (16ms verification + 90ms entity extraction on A100), compared to 0.85 F1 for prompt-based filtering and 0.65 for rule-based DLP. To empirically validate the need for external enforcement, we evaluate eight frontier LLMs in execution-oriented multi-agent workflows with per-agent domain world models. All models exhibit substantial violation rates (14-98%), with cross-domain data flows showing systematically higher violation rates than same-domain flows. These results indicate that self-avoidance is unreliable and that multi-agent security benefits from a centralized enforcement layer operating above individual agents.

Malware development and detection have undergone significant changes in recent years as modern concepts, such as machine learning, have been used for both adversarial attacks and defense. Despite intensive research on Windows Portable Executable (PE) files, there is minimal work on Linux Executable and Linkable Format (ELF). In this work, we summarize the academic papers submitted in this field and develop a new adversarial malware generator for the ELF format. Using a variety of metrics, we thoroughly evaluated our generator and achieved an Evasion Rate of 67.74 % while changing the confidence of the malware detector by -0.50 in the mean case for the dataset used. In our approach, we chose MalConv as the target classifier. Using this classifier, we found that the most successful modifications used strings typical of benign files as a data source. We conducted a variety of experiments and concluded that the target classifier appears sensitive to strings at any location within the executable file.

This work proposes a structural approach to concept drift detection in malware classification using decision tree rulesets. Classifiers are trained across temporal windows on the EMBER2024 dataset, and drift is quantified by comparing extracted rule representations using feature importance, prediction agreement, activation stability, and coverage metrics. These metrics are correlated with both accuracy degradation and data distribution shift as complementary drift indicators. The approach is evaluated across six malware families using fixed-interval and clustering-based windowing in family-vs-benign and family-vs-family settings, and compared against RIPPER and Transcendent baselines. Results show that fixed two-month windowing with feature-level Pearson correlation is the most reliable configuration, being the only one where all family pairs produce positive drift-accuracy correlations. The methods are complementary - no single approach dominates across all pairs.

Blockchain wallets conventionally follow an ownership model where possession of a private key grants unilateral control. However, this assumption is brittle for emerging settings such as AI agent wallets, organizational custody, and enterprise payroll, where multiple actors must coordinate without exposing secrets or leaking internal activity. We present PASS, a Provenanced Access Subaccount System that replaces role-based or identity-based control with provenance-based control: assets can only be used by subaccounts that can trace custody back to a valid deposit. A simple Inbox-Outbox mechanism ensures all external actions have verifiable lineage, while internal transfers remain private and indistinguishable from ordinary EOAs. We formalize PASS in Lean 4 and prove core invariants, including privacy of internal transfers, asset accessibility, and provenance integrity. We implement a prototype with enclave backends on AWS Nitro Enclaves and dstack Intel TDX, integrate with WalletConnect, and benchmark throughput across wallet operations. These results show that provenance-based wallets are both implementable and efficient. PASS bridges today's gap between strict self-custody and flexible shared access, advancing the design space for practical, privacy-preserving custody.

Machine learning-based malware detectors are increasingly vulnerable to adversarial examples. Traditional defenses, such as one-shot adversarial training, often fail against adaptive attackers who use reinforcement learning to bypass detection. This paper proposes a robust defense framework based on bilevel optimization, explicitly modeling the strategic interaction between a defender and an attacker as an adversarial co-evolutionary process. We evaluate our approach using the MAB-malware framework against three distinct malware families: Mokes, Strab, and DCRat. Our experimental results demonstrate that while standard classifiers and basic adversarial retraining often remain vulnerable, showing evasion rates as high as 90 %, the proposed bilevel optimization approach consistently achieves near-total immunity, reducing evasion rates to 0 - 1.89 %. Furthermore, the iterative framework significantly increases the attacker's query complexity, raising the average cost of successful evasion by up to two orders of magnitude. These findings suggest that modeling the iterative cycle of attack and defense through bilevel optimization is essential for developing resilient malware detection systems capable of withstanding evolving adversarial threats.

Self-supervised learning (SSL) encoders are invaluable intellectual property (IP). However, no existing SSL watermarking for IP protection can concurrently satisfy the following two practical requirements: (1) provide ownership verification capability under black-box suspect model access once the stolen encoders are used in downstream tasks; (2) be robust under adversarial watermark detection or removal, because the watermark samples form a distinguishable out-of-distribution (OOD) cluster. We propose ArmSSL, an SSL watermarking framework that assures black-box verifiability and adversarial robustness while preserving utility. For verification, we introduce paired discrepancy enlargement, enforcing feature-space orthogonality between the clean and its watermark counterpart to produce a reliable verification signal in black-box against the suspect model. For adversarial robustness, ArmSSL integrates latent representation entanglement and distribution alignment to suppress the OOD clustering. The former entangles watermark representations with clean representations (i.e., from non-source-class) to avoid forming a dense cluster of watermark samples, while the latter minimizes the distributional discrepancy between watermark and clean representations, thereby disguising watermark samples as natural in-distribution data. For utility, a reference-guided watermark tuning strategy is designed to allow the watermark to be learned as a small side task without affecting the main task by aligning the watermarked encoder's outputs with those of the original clean encoder on normal data. Extensive experiments across five mainstream SSL frameworks and nine benchmark datasets, along with end-to-end comparisons with SOTAs, demonstrate that ArmSSL achieves superior ownership verification, negligible utility degradation, and strong robustness against various adversarial detection and removal.

Private Information Retrieval (PIR) allows clients to retrieve database entries without leaking retrieval indices, yet malicious servers seriously compromise retrieval correctness. Existing Authenticated PIR (APIR) schemes resist selective-failure attacks but rely on computational hardness assumptions. In contrast, information-theoretic PIR with Result Verification (itPIR-RV) achieves integrity without computational assumptions, yet only provides relaxed query privacy with no defense against selective-failure attacks. This paper focuses on unconditionally secure information-theoretic APIR (itAPIR) constructions. We propose the rigorous information-theoretic security definition for itAPIR with statistical privacy against selective-failure attacks and integrity as core properties, formalize the hierarchical relation between itAPIR and itPIR-RV as a relaxed variant with identical integrity but basic query privacy, and prove a conversion theorem that valid itPIR-RV schemes can be directly upgraded to secure itAPIR with no extra overhead. Our work bridges the theoretical gap, simplifies itAPIR design, and enables quantum-resistant PIR in malicious server environments.

Watermarking has emerged as a promising technique for tracing the authorship of content generated by large language models (LLMs). Among existing approaches, the KGW scheme is particularly attractive due to its versatility, efficiency, and effectiveness in natural language generation. However, KGW's effectiveness degrades significantly under low-entropy settings such as code generation and mathematical reasoning. A crucial step in the KGW method is random vocabulary partitioning, which enables adjustments to token selection based on specific preferences. Our study revealed that the next-token probability distribution plays an critical role in determining how much, or even whether, we can modify token selection and, consequently, the effectiveness of watermarking. We refer to this characteristic, associated with the probability distribution of each token prediction, as watermark strength. In cases of random vocabulary partitioning, the lower bound of watermark strength is dictated by the next-token probability distribution. However, we found that, by redesigning the vocabulary partitioning algorithm, we can potentially raise this lower bound. In this paper, we propose SSG (Sort-then-Split by Groups), a method that partitions the vocabulary into two logit-balanced subsets. This design lifts the lower bound of watermark strength for each token prediction, thereby improving watermark detectability. Experiments on code generation and mathematical reasoning datasets demonstrate the effectiveness of SSG.

Scalar multiplication kP is the operation most frequently targeted in Elliptic Curve (EC) cryptosystems. To protect against single-trace Side-Channel Analysis (SCA) attacks, the atomicity principle and various atomic block patterns have been proposed in the past. In this work we use our software and hardware implementations to demonstrate that binary right-to left and left-to-right kP algorithms, when implemented with Chevallier-Mames atomic block patterns, are still vulnerable to single-trace SCA attacks. The vulnerability remains true for the left-to-right kP algorithm with projective coordinate randomization.

The offensive security landscape is highly fragmented: enterprise platforms avoid memory-corruption vulnerabilities due to Denial of Service (DoS) risks, Automatic Exploit Generation (AEG) systems suffer from semantic blindness, and Large Language Model (LLM) agents face safety alignment filters and "Live Fire" execution hazards. We introduce Automation-Exploit, a fully autonomous Multi-Agent System (MAS) framework designed for adaptive offensive security in complex black-box scenarios. It bridges the abstraction gap between reconnaissance and exploitation by autonomously exfiltrating executables and contextual intelligence across multiple protocols, using this data to fuel both logical and binary attack chains. The framework introduces an adaptive safety architecture to mitigate DoS risks. While it natively resolves logical and web-based vulnerabilities, it employs a conditional isomorphic validation for high-risk memory-corruption flaws: if the target binary is successfully exfiltrated, it dynamically instantiates a cross-platform digital twin. By enforcing strict state synchronization, including libc alignment and runtime file descriptor hooking, potentially destructive payloads are iteratively debugged in an isolated replica. This enables a highly risk-mitigated "one-shot" execution on the physical target. Empirical evaluations across eight scenarios, including undocumented zero-day environments to rule out LLM data contamination, validate the framework's architectural resilience, demonstrating its ability to prevent "live fire" crashes and execute risk-mitigated compromises on actual targets.

A growing number of Internet of Things (IoT) devices are used across consumer, medical, and industrial domains. They interact with their environment through sensors and actuators and connect to networks such as the Internet. Because sensors may collect sensitive data and actuators can trigger physical actions, security, privacy, and safety are major challenges. Threat modelling can help identify risks, but established IT-focused methods transfer to the IoT only to a limited extent. In this paper, a new modelling technique specifically for IoT devices called Cyber-Physical Data Flow Diagram (CPDFD) is proposed that also allows modelling of hardware with the aim to support manufacturers in identifying threats and developing countermeasures. The technique was examined through an experimental study and a survey with interviews. The results suggest that numerous other attack scenarios can be found through the modelling technique, improving the identification of threats to IoT devices.

This paper proposes a resource-aware allocation model for layered intrusion detection in het erogeneous networks. Monitoring traffic at higher protocol layers improves the ability to detect sophisticated attacks, but it also increases computational and storage costs. The problem is formu lated as an integer linear program that assigns a single monitoring depth, ranging from Ethernet to the application layer, to each device, while accounting for device importance, attack probability, layer-dependent detection rates, and per-layer monitoring costs. The model further enforces a global resource budget, a minimum monitoring level for critical devices, and maximum-feasibility limits for constrained devices such as simple IoT sensors. The formulation is solved with the SCIP optimization framework on a small heterogeneous network of six devices, and the resulting allocation illustrates how the model concentrates monitoring effort on important and high-risk devices while respecting feasibility and budget constraints.

The widespread availability of large-scale code datasets has accelerated the development of code large language models (CodeLLMs), raising concerns about unauthorized dataset usage. Dataset poisoning offers a proactive defense by reducing the utility of such unauthorized training. However, existing poisoning methods often require full dataset poisoning and introduce transformations that break code compilability. In this paper, we introduce FunPoison, a functionality-preserving poisoning approach that injects short, compilable weak-use fragments into executed code paths. FunPoison leverages reusable statement-level templates with automatic repair and conservative safety checking to ensure side-effect freedom, while a type-aware synthesis module suppresses static analysis warnings and enhances stealth. Extensive experiments show that FunPoison achieves effective poisoning by contaminating only 10% of the dataset, while maintaining 100% compilability and functional correctness, and remains robust against various advanced code sanitization techniques.

In agentic workflows, LLMs frequently process retrieved contexts that are legally protected from further training. However, auditors currently lack a reliable way to verify if a provider has violated the terms of service by incorporating these data into post-training, especially through Reinforcement Learning (RL). While standard auditing relies on verbatim memorization and membership inference, these methods are ineffective for RL-trained models, as RL primarily influences a model's behavioral style rather than the retention of specific facts. To bridge this gap, we introduce Behavioral Canaries, a new auditing mechanism for RLFT pipelines. The framework instruments preference data by pairing document triggers with feedback that rewards a distinctive stylistic response, inducing a latent trigger-conditioned preference if such data are used in training. Empirical results show that these behavioral signals enable detection of unauthorized document-conditioned training, achieving a 67% detection rate at a 10% false-positive rate (AUROC = 0.756) at a 1% canary injection rate. More broadly, our results establish behavioral canaries as a new auditing mechanism for RLFT pipelines, enabling auditors to test for training-time influence even when such influence manifests as distributional behavioral change rather than memorization. We release our code at: https://github.com/CRChenCode/behavioral_canary.

Accurate mapping between Common Vulnerabilities and Exposures (CVE) and Common Weakness Enumeration (CWE) entries is critical for effective vulnerability management and risk assessment. However, public databases, such as the National Vulnerability Database (NVD), suffer from inconsistent and incomplete CVE to CWE mappings, complicating automated analysis and remediation. We introduce FixV2W, a lightweight approach that leverages knowledge graph embeddings and longitudinal trends to improve mapping accuracy of the NVD. FixV2W systematically analyzes historical remapping patterns and leverages hierarchical relationships within NVD and CWE data to predict more precise CWE mappings for vulnerabilities linked to Prohibited or Discouraged categories. We run extensive experimental evaluation of FixV2W, based on test data set collected between August 2021 and December 2024. Considering the Top 10 ranked predictions, the results show that FixV2W predicts the correct CWE mappings for 69% of exploited vulnerabilities that had invalid CWEs before they were exploited. We also show that FixV2W significantly improves the performance of ML models relying on NVD data. For instance, for a model geared at uncovering unknown CVE-CWE mappings, FixV2W improves the Mean Reciprocal Rank (MRR) from 0.174 to 0.608. These results show that FixV2W is a promising approach to identify and thwart emerging threats.

Existing research typically treats privacy policies as flat, uniform text, extracting information without regard for the document's logical hierarchy. Disregard for structural cues of section headings designed to guide the reader, often leads automated methods to entangle distinct data practices, particularly when linking sensitive data items to their specific purposes. To address this, we introduce PrivSTRUCT, a novel and systematic encoder and decoder combined framework that to untangle complex privacy disclosures. Benchmarking against the state-of-the-art tool PoliGrapher reveals that PrivSTRUCT robustly extracts more than x2 the number of data item and purpose excerpts while retaining developer-defined structural cues. By applying PrivSTRUCT to a large-scale dataset of 3,756 Android apps, we uncover a critical transparency gap: the probability of developers overstating a data purpose is 20.4% higher for first-party collection and 9.7% higher for third-party sharing when they rely on globally defined purposes rather than specific, locally scoped disclosures. Alarmingly, we find that sensitive third-party data flows such as sharing financial data for analytics are frequently diluted and entangled into generic or unrelated categories, highlighting a persistent failure in the current purpose disclosure landscape.

Large language model (LLM) agents increasingly issue API calls that mutate real systems, yet many current architectures pass stochastic model outputs directly to execution layers. We argue that this coupling creates a safety risk because model correctness, context awareness, and alignment cannot be assumed at execution time. We introduce Sovereign Agentic Loops (SAL), a control-plane architecture in which models emit structured intents with justifications, and the control plane validates those intents against true system state and policy before execution. SAL combines an obfuscation membrane, which limits model access to identity-sensitive state, with a cryptographically linked Evidence Chain for auditability and replay. We formalize SAL and show that, under the stated assumptions, it provides policy-bounded execution, identity isolation, and deterministic replay. In an OpenKedge prototype for cloud infrastructure, SAL blocks 93% of unsafe intents at the policy layer, rejects the remaining 7% via consistency checks, prevents unsafe executions in our benchmark, and adds 12.4 ms median latency.

This study considers the task of applying artificial intelligence to recognize bank fraud. In recent years, due to the COVID19 pandemic, bank fraud has become even more common due to the massive transition of many operations to online platforms and the creation of many charitable funds that criminals can use to deceive users. The present work focuses on machine learning algorithms as a tool well suited for analyzing and recognizing online banking transactions. The study`s scientific novelty is the development of machine learning models for identifying fraudulent banking transactions and techniques for preprocessing bank data for further comparison and selection of the best results. This paper also details various methods for improving detection accuracy, i.e., handling highly imbalanced datasets, feature transformation, and feature engineering. The proposed model, which is based on an artificial neural network, effectively improves the accuracy of fraudulent transaction detection. The results of the different algorithms are visualized, and the logistic regression algorithm performs the best, with an output AUC value of approximately 0,946. The stacked generalization shows a better AUC of 0.954. The recognition of banking fraud using artificial intelligence algorithms is a topical issue in our digital society.

The rapid growth of Internet of Things (IoT) devices has led to large-scale continuous data streams that require realtime processing. Traditional cloud-centric architectures fail to meet low-latency and bandwidth efficiency requirements due to network delays and high data transmission overhead. This paper proposes EdgeStream, a lightweight edge-based framework for real-time streaming analytics in IoT environments. The system integrates edge nodes for local processing with a cloud backend for coordination and storage, using MQTT-based communication and distributed processing with anomaly detection. Analytical models for latency, throughput, and bandwidth are developed to evaluate performance. Experimental results, compared to cloudbased systems, show up to 92.8

Automated red-teaming methods for large language models typically optimize attack prompts within a fixed, human-designed strategy, leaving the attack strategy itself unchanged. We instead optimize the strategy. We propose AutoRISE, a method that searches over executable attack programs rather than individual prompts. At each iteration, a coding agent edits a strategy and a fixed evaluation harness scores the resulting attacks, returning both a scalar objective and per-example diagnostics that guide subsequent edits. This allows structural changes, including new attack components and altered control flow, that prompt-level methods do not directly express. We also release two benchmark suites developed on disjoint target sets and evaluate on 11 models from five families against seven established jailbreak datasets. Across held-out models, AutoRISE improves average attack success rate by 17.0 points over the strongest baseline, and improves attack success by up to 16 points on frontier targets with low baseline success rates. Ablations against parametric and strategy-library baselines suggest that these gains arise from unrestricted program search, particularly compositional techniques and control-flow edits. AutoRISE operates in a black-box, inference-only setting, requiring no fine-tuning, human annotation, or GPU compute.

This article applies postphenomenological theory to the field of cybersecurity risk management, arguing that formal risk models function as mediating artifacts that shape how security practitioners or analysts perceive, interpret, and act on threats. Based on Don Ihde's taxonomy on human-technology relationships and Peter-Paul Verbeek's extended mediational framework, the Contextual and Multimodal Hazard Impact Index (CIIM), an original dynamic risk model presented as an empirical case study, is analyzed. CIIM is formally defined as CIIM(t+1) = [A T(t) V(t) E(t)] / R(t) + {alpha} P(t), where the condition R(t) 0 is not treated as a computational artifact to be smoothed out, but as a genuine systemic collapse that signals singularity. This design choice constitutes a deliberate phenomenological move, allowing organizational fragility to be made visible in a way that previous CVSS-based and probabilistic models conceal. In addition, we examine how CIIM's time projection (t+1) and its hybrid machine learning architecture, combining LSTM/GRU, XGBoost, and Reinforcement Learning, produce a new form of technological intentionality that structures practitioner or analyst attention and ethical deliberation. The article concludes by establishing implications for the ethical design of cybersecurity instrumentation and for the post-phenomenological methodology itself, proposing the concept of 'phenomenology of collapse' as a contribution to the empirical philosophy of technology.

In enterprise fraud detection, model accuracy alone is insufficient when insiders can tamper with audit logs or bypass approval workflows. Real-world incidents show that fraud often persists not because detection algorithms fail, but because the audit trail itself is controllable by privileged operators. This exposes a fundamental trust gap: *who audits the auditor?* We present a tamper-evident fraud detection system that anchors both ML predictions and workflow execution to an immutable blockchain ledger. Rather than using blockchain as passive storage, we enforce the entire approval process through smart contracts, ensuring that every transaction, prediction, and explanation is atomically recorded and cannot be retroactively modified. Our detection module achieves competitive accuracy (F1 = 0.895, PR-AUC = 0.974) while providing cryptographically verifiable decision trails that support regulatory auditability requirements (e.g., GDPR Article 22). System evaluation shows sub-25 ms inference latency and economically viable deployment on Layer-2 networks at under \$0.01 per transaction (validated against PolygonScan data), supporting enterprise-scale workloads of 10,000+ monthly payments.

Security Operations Centers (SOCs) are pivotal in modern enterprises. Tasked to monitor complex network environments constantly under attack, SOCs can be active 24/7 and can include hundreds of operators supported by state-of-the-art technologies. Abundant research has studied the internal processes of SOCs, highlighting their pros and cons, as well as the challenges faced by SOC analysts – such as dealing with the overwhelming number of false alarms triggered by automated security mechanisms. In this context, we wonder: given that "someone" must triage the alarms, and that such triaging must be grounded on established knowledge or evidence-based reasoning, can SOC employees justify why a certain decision was taken while triaging alarms? Answering such a research question (RQ) can better guide future efforts. We hence tackle this RQs. First, via a systematic literature review across 257 research documents, we provide evidence that such RQ received limited attention so far. Then, we partner-up with a real-world SOC and carry out a field study (n=12) with SOC employees. We show them real alarms raised in their SOC, and inquire whether such alarms are indicative of true security problems or not. Then, we ask to explain their decision. We found that while most analysts were able to separate "true from false" alarms (the decision was correct in 83% of the cases), a correct justification was hardly provided (only 39% of the provided explanations reflected the actual root cause). Ultimately, our results highlight the need for decision-support systems that help SOC analysts not only make the right call – but also understand and articulate why it is right.

We present CrossCommitVuln-Bench, a curated benchmark of 15 real-world Python vulnerabilities (CVEs) in which the exploitable condition was introduced across multiple commits - each individually benign to per-commit static analysis - but collectively critical. We manually annotate each CVE with its contributing commit chain, a structured rationale for why each commit evades per-commit analysis, and baseline evaluations using Semgrep and Bandit in both per-commit and cumulative scanning modes. Our central finding: the per-commit detection rate (CCDR) is 13% across all 15 vulnerabilities - 87% of chains are invisible to per-commit SAST. Critically, both per-commit detections are qualitatively poor: one occurs on commits framed as security fixes (where developers suppress the alert), and the other detects only the minor hardcoded-key component while completely missing the primary vulnerability (200+ unprotected API endpoints). Even in cumulative mode (full codebase present), the detection rate is only 27%, confirming that snapshot-based SAST tools often miss vulnerabilities whose introduction spans multiple commits. The dataset, annotation schema, evaluation scripts, and reproducible baselines are released under open-source licenses to support research on cross-commit vulnerability detection.

Large language models (LLMs) are increasingly integrated into sensitive workflows, raising the stakes for adversarial robustness and safety. This paper introduces Transient Turn Injection(TTI), a new multi-turn attack technique that systematically exploits stateless moderation by distributing adversarial intent across isolated interactions. TTI leverages automated attacker agents powered by large language models to iteratively test and evade policy enforcement in both commercial and open-source LLMs, marking a departure from conventional jailbreak approaches that typically depend on maintaining persistent conversational context. Our extensive evaluation across state-of-the-art models-including those from OpenAI, Anthropic, Google Gemini, Meta, and prominent open-source alternatives-uncovers significant variations in resilience to TTI attacks, with only select architectures exhibiting substantial inherent robustness. Our automated blackbox evaluation framework also uncovers previously unknown model specific vulnerabilities and attack surface patterns, especially within medical and high stakes domains. We further compare TTI against established adversarial prompting methods and detail practical mitigation strategies, such as session level context aggregation and deep alignment approaches. Our study underscores the urgent need for holistic, context aware defenses and continuous adversarial testing to future proof LLM deployments against evolving multi-turn threats.

Autonomous Vehicles (AVs) increasingly depend on Multi-Sensor Fusion (MSF) to combine complementary modalities such as cameras and LiDAR for robust perception. While this redundancy is intended to safeguard against single-sensor failures, the fusion process itself introduces a subtle and underexplored vulnerability. In this work, we investigate whether an attacker can bypass MSF's redundancy by fabricating cross-sensor consistency, making multiple sensors agree on the same false object. We design a coordinated, data-level (early-fusion) attack that emulates the outcome of two synchronized physical spoofing sources: an infrared (IR) projection that induces a false camera detection and a LiDAR signal injection that produces a matching 3D point cluster. Rather than implementing the physical attack hardware, we simulate its sensor-level outcomes by inserting perspective-aware image patches and synthetic LiDAR point clusters aligned in 3D space. This approach preserves the perceptual effects that real IR and IEMI-based spoofing would create at the sensor output. Using 400 KITTI scenes, our large-scale evaluation shows that the coordinated spoofing deceives a state-of-the-art perception model with an 85.5% successful attack rate. These findings provide the first quantitative evidence that malicious cross-modal consistency can compromise MSF-based perception, revealing a critical vulnerability in the core data-fusion logic of modern autonomous vehicle systems.