Papers

Money mule accounts are critical facilitators of financial fraud, yet detecting them at scale remains challenging due to the heterogeneous nature of transactional and behavioural data. We present an end-to-end pipeline for customer-level mule detection comprising three stages: (1) a LightGBM classifier trained on 280 engineered features spanning transaction patterns, account demographics, network topology, and temporal behaviour; (2) a TreeSHAP attribution layer that decomposes each prediction into feature contributions; and (3) a large language model (LLM) module that converts SHAP attributions into analyst-facing natural-language narratives. We evaluate across three open-weight LLM families and assess explanation quality through analyst feedback. In a live production deployment, the system achieves a yield rate of 89%, up from 61% under the incumbent rule-based system, with monthly alert volume expanding from 211 to 302, reflecting broader true-positive coverage rather than increased noise. This corresponds to a 60% incremental adverse detection beyond existing review workflows, substantially outperforming the rule-based approach. Qualitative feedback from analysts indicates that LLM-generated narratives reduce cognitive load during alert triage. We further discuss implications of deploying LLM-augmented explainability in regulated financial environments.

The increasing number of online banking and financial services on public-facing networks has caused security of cryptographic protocols to become a serious, systemic problem. RSA-2048 is the key length that is used in the majority of transactions that travel over the Internet every day. A very insidious threat vector, the Harvest Now, Decrypt Later (HNDL) paradigm, is already in existence with the adversarial nation-state actors and well-funded threat actors. The exposure window is not a time in the future it is the NOW moment! Quantum computation is a certain threat to the current public-key cryptographic infrastructure, and one that is moving ever closer. In this paper, the authors take up both of these challenges, designing, implementing and testing an integrated system based on a Dual-Wrap Hybrid Key Encapsulation Mechanism (DW-HKEM) and a live cryptographic analytics dashboard. Both the RSA-256 and ML-KEM-768 are hosted within the DW-HKEM framework and an AES-256 session key is wrapped in each of them, with a composite key created using a Key Derivation Function based on SHA-256, resulting in combinatorial security against classical and quantum adversaries and backwards compatibility with existing banking systems. The CryptoX real-time dashboard offers time tracking for each operation, session-level analytics and visualization of performance trends for immediate use in enterprise banking security operations. The results from empirical benchmarking across 100 independent iterations show a total overhead of about 1.58ms, which is well within the enterprise deployment threshold, proving the architecture viable as a practical, scalable reference for the financial-sector's cryptographic infrastructure and applications to move to a post-quantum era. We also provide complete source code of the work carried out.

Modern machine learning (ML) pipelines depend heavily on third party libraries for graph compilation and hardware acceleration. While current practices audit data and model artifacts or rely on file integrity checks, the execution environment remains implicitly trusted. This blind spot enables active threats where a malicious runtime module interacts directly with live training and inference dynamics: exploiting this interaction allows the Trojan to support complex objectives that are challenging for static code or binary modifications, achieving manipulations impossible for standard data and model level attacks. We expose this vulnerability by presenting AiSPY, a parasitic infrastructure Trojan that subverts MLsystems through an active observe and execute paradigm. Operating within the computation graph, AiSPY monitors transient tensor states to perform targeted, stealthy manipulations with negligible overhead. To violate confidentiality, the Trojan identifies all critical training hyperparameters and covertly exfiltrates them via model weights or output logits. To break integrity, it acts as a gradient amplifier: by observing steganographic triggers, it transforms other- wise weak data poisoning into effective backdoor attacks, increasing success rates from near zero to 100%. We further demonstrate broad extensibility across the machine learning lifecycle by validating auxiliary attacks in the appendix, including subpopulation label flipping, availability disruptions, and inference stage manipulations. Importantly, the evaluated malware scanners do not flag AiSPY because current public rule sets lack coverage for ML runtime Trojans, while the associated poisoned inputs and resulting compromised models bypass state-of-the-art inspection tools. We demonstrate the practicality of this threat with an implementation in the ONNX Runtime training and inference engines.

Agentic retrieval-augmented generation (RAG) systems increasingly retrieve external evidence and orchestrate tools for knowledge-intensive applications. In Multi-Hop question answering, agents chain facts across documents. Existing defenses focus on content poisoning, which injects false facts, and prompt injection, which embeds directives. We identify a third attack surface: the salience channel, through which fact position, emphasis, framing, and semantic proximity can redirect reasoning even when all retrieved claims are true and no instructions are present. We formalize Salience Induction as truth-preserving edits that redirect Multi-Hop attribute binding while leaving the retrieval trace semantically intact. We define six Salience-Editing operator classes and build an iterative proposer-verifier pipeline under factual and stealth constraints. We also introduce SalientWiki-MH, a decoy-annotated Multi-Hop benchmark. Evaluations across five frontier model families (GPT, Claude, Gemini, DeepSeek, and Qwen) and three agent architectures (ReAct, Reflexion, and tool-calling) show broad generalization. Under a 30% edit budget, Salience Induction achieves an 83.3% attack success rate; the strongest evaluated baseline defense leaves 75.7% post-defense ASR. Untargeted rewriting further reduces attacks only by degrading neutral task success. Our lightweight input-side defense, Salience Normalization, reduces attack success to 15.3% under standard attacks and 23.6% under an adaptive attack. These results show that truthfulness and instruction filtering alone are insufficient: robust agentic RAG also requires defenses against salience-relevance decoupling.

Modern cloud platforms increasingly combine strong software isolation mechanisms with shared hardware resources to improve performance and resource efficiency. Conventional containers do this by sharing the host kernel directly, whereas sandboxed runtimes (e.g., gVisor) and VM-based runtimes (e.g., Kata, QEMU/KVM) provide progressively stronger isolation. In all cases, when tenants access host-backed filesystem state, the host page cache can remain shared and observable. Although OS-managed, this page-cache channel forms an OS-mediated microarchitectural timing side channel whose signal is shaped by the processor microarchitecture, memory and storage hierarchies, and virtualization mechanisms. We thus investigate whether unprivileged timing measurements can reveal page-cache residency across these isolation boundaries. Our evaluation covers Docker; gVisor with systrap and KVM; Kata Containers using QEMU and Cloud Hypervisor with shared host filesystems; Kata using QEMU, Cloud Hypervisor, and Firecracker with block-device-backed storage; and QEMU/KVM virtual machines under multiple host cache policies. Our results show that the timing signal persists whenever the I/O path exposes shared, host-cacheable file-backed objects, including under OverlayFS layers, virtio-fs exports, and loop-backed block devices. However, direct I/O and dedicated block devices substantially attenuate or eliminate the signal. Virtualization therefore reshapes leakage through added latency and algorithmic noise but does not remove the underlying dependence on shared hardware and cache state. We showcase this through a case study in which we recover coarse-grained activity from a WordPress deployment backed by MySQL. These results place page-cache attacks within the broader class of OS-mediated microarchitectural timing channels and motivate coordinated hardware, virtualization, and OS support for timing isolation.

Model hosting hubs (e.g., Hugging Face) are vulnerable to supply chain attacks that enable remote code execution on trusted user environments. Attackers often distribute malicious Pre-trained ML models (PTMs) via model hubs. In this paper, we present novel attacks against PTMs and model hubs called SHADOWPICKLE. SHADOWPICKLE includes three (3) stealthy pickle deserialization attacks that enable malicious behaviors and evade state-of-the-art (SOTA) model scanners. These attacks leverage the external module import mechanism of the Pickle Virtual Machine (VM) to execute malicious payloads during deserialization. Additionally, we provide PICKLEBENCH, a dynamic and extensible benchmark for automatically injecting SHADOWPICKLE into arbitrary benign PTM models. Our evaluation shows that SHADOWPICKLE evades ten SOTA scanners, and four model hubs. SHADOWPICKLE (Overwritten) has a 63% evasion rate across scanners, and up to 50% higher evasion rates than existing attacks. Besides, PICKLEBENCH is up to 25.6% more challenging than three SOTA benchmarks. Finally, we provide security recommendations for mitigating our attacks and improving the effectiveness of existing scanners. Our findings highlight the limitations of existing PTM scanners and suggest directions for improvements.

Non-interference properties, spanning confidentiality and integrity, have long enjoyed a position as the high water mark of program security guarantees. Information flow type systems comprise the primary means for obtaining non-interference properties of programs, but their potential as a holy grail for secure programming has remained latent. Prior work bifurcates the type system along confidentiality and integrity, resulting in duplicate reasoning machinery and complex specifications. Furthermore, long-held wisdom dictates that non-interference must be weakened with downgrading mechanisms to accommodate the needs of practical programs, nearly all of which violate confidentiality and integrity in the course of fulfilling their purpose. This often pierces abstraction barriers and compromises modular reasoning. We introduce parametric information flow, which uses recent insights from modal type theory to shed light on these issues. In particular, we draw inspiration from work on the open and closed modalities, highlighting their rich interplay. Though each individual modality finds uses throughout the literature, our key insight is that their joint interaction suffices to reconstruct full-spectrum information flow reasoning, producing a single framework accounting for both confidentiality and integrity. Downgrading and analogues of advanced reasoning tools in the lineage of robust declassification are recovered without extensions to our theory, strengthening prior results. We show non-interference via a binary logical relations argument, realizing robustness as an ordinary 2-hyperproperty mediated by our modalities. Our work reveals state-of-the-art downgrading mechanisms to be wholly compatible with those for abstraction and modularity, arising precisely from the semantics of the latter under full-strength non-interference.

Autonomous drone-based services have been gaining significant interest across various application domains due to their mobility, flexibility, cost-effectiveness, and capability to integrate various sensors and actuators. However, operational failures or cyberattacks targeting drone systems can lead to severe economic impacts and safety concerns. Hence, ensuring secure and reliable autonomous drone operations is essential for the safe deployment of drone-enabled services. Nevertheless, the traditional security measures fall short due to drones' limited computational resources and power budget (battery), as well as the temporal and sequential behavior of drone operations due to drones' mobile nature. In this paper, we address this gap by utilizing Recurrent Neural Networks (RNNs), specifically focusing on Long Short-Term Memory (LSTM) networks, for autonomous drone operation security and reliability due to their temporal and sequential analysis capabilities. We leverage these capabilities for anomaly detection in autonomous drone sensor data and operation commands, which we refer to as Denial of Usage Detection Engine IDS (DUDE-IDS). We integrated the proposed DUDE-IDS into the drone mission computer (i.e., operating directly on drones rather than an edge node or ground control stations) to monitor data flows and to detect potential threats in real-time. Extensive experimental results demonstrate the effectiveness of this approach in identifying anomalies associated with GPS spoofing, Man-in-the-Middle (MITM), replay, and Denial-of-Service (DoS) attacks with 98% accuracy. We also evaluate our resource utilization and power consumption under different configurations, demonstrating the applicability of our approach in active drone operations in real-time.

Online scams continue to cause substantial financial and personal harm. As a result, detection systems based on Large Language Models (LLMs) have been integrated into security products ranging from email gateways and browser extensions to fraud-monitoring dashboards. As this adoption accelerates, a common belief has taken hold: that these models are broadly suitable for scam detection. In this work, we investigate whether LLMs, with their strong capabilities in understanding intent, context, and reasoning, can effectively detect scams across diverse scenarios without task-specific fine-tuning. We curate and release a unique benchmark dataset of real-world scams spanning multiple formats and topics. We evaluate nine LLMs of varying sizes and architectures, examining their performance under different prompting strategies and comparing them to a fine-tuned BERT-based classifier. Our results show that while larger LLMs generally outperform smaller ones, effective prompting substantially boosts the performance of smaller models. Moreover, LLMs are better at generalizing to unseen scams compared to fine-tuned models, suggesting that pre-trained knowledge contributes meaningfully to scam detection. We release our dataset and evaluation framework to facilitate future research in robust scam detection using language models.

Traditional time-lock puzzles enforce delayed access to encrypted secrets by relying on inherently sequential computational steps. However, since software-based constructions impose no fundamental bound on the physical execution speed of individual steps, they remain vulnerable to hardware acceleration and improved implementations. In practice, this renders existing schemes “step-lock” rather than true “time-lock” puzzles, making long-term delay guarantees highly speculative against decades of unpredictable hardware advancement. To address this fundamental limitation, we introduce the Physical Time-Lock Puzzle (P-TLP), a new paradigm that anchors solving delay directly to the intrinsic, hardware-bounded evaluation latency of silicon hardware. By leveraging noisy Physical Unclonable Functions (PUFs) as non-parallelizable delay oracles, P-TLPs establish resistance against both parallel computation and algorithmic acceleration under the random oracle model. To tolerate PUF evaluation noise while maintaining a tightly concentrated and predictable solving-delay window, we propose a composite puzzle architecture that combines multiple independent basic puzzles. We formally prove the optimality of a sequential greedy solving strategy and derive tightly concentrated solving-delay windows using Hoeffding and Berry-Esseen bounds. We additionally prove early-solve hardness, establishing that adversaries with substantial classical computation budgets cannot bypass the physically-enforced PUF evaluation bottleneck. We validate our theoretical analysis with an FPGA prototype built around a configurable Ring Oscillator PUF, confirming that our analytical delay predictions tightly match empirical measurements. Our results demonstrate that P-TLPs can deliver highly predictable, long-term delay guarantees, enabling practical deployment in high-value applications such as digital legacy management.

We model cryptographic auditing of off-chain data as a Constrained MDP (CMDP) under partial observability: the storage node's hidden type and corruption state make the problem a POMDP, while a miss-rate ceiling rho imposes an explicit security constraint. We propose DRQN-CMDP, a Deep Recurrent Q-Network whose GRU layer maintains a belief over the latent node type, paired with Lagrangian dual ascent that adapts the miss-rate penalty lambda automatically. A pairing-free homomorphic-MAC primitive supplies O(1) on-chain verification cost. Across 13 methods–four DQN variants, PPO, A2C, PPO-Lagrangian, a stateful Bayesian heuristic, three fixed-rule baselines, and an oracle-informed heuristic–DRQN-CMDP achieves a favourable balance: 83% lower gas than fixed high-frequency auditing, single-digit miss rate (7.5%), and moderate detection latency–a combination no other method matches across all three objectives simultaneously.

Recently, text-to-video (T2V) models have been widely deployed, sparking growing concerns over their robustness against jailbreak attacks. Existing jailbreak methods, mostly adapted from text-to-image attacks, suffer notable drawbacks when applied to T2V systems. They fail to fully leverage temporal consistency, an inherent characteristic of video generation. Besides, these methods demand heavy video query optimization, which is infeasible in practical black-box scenarios. Their adversarial prompt search is also driven by heuristic local signals, lacking principled structured exploration strategies. To tackle these limitations, we propose BSB, a structured, query-efficient jailbreak framework for T2V models. BSB harnesses temporal consistency by encoding harmful intent as the transition between two individually harmless boundary states. Under this paradigm, the attack targets boundary-state pairs whose interpolation tends to produce unsafe intermediate frames during video generation. Directly evaluating all candidate pairs within the video space incurs prohibitive computation cost. Instead, BSB conducts Monte Carlo Tree Search (MCTS) in a cheaper textual proxy space and regularly calibrates search outcomes with sparse video-level evaluations. We conduct comprehensive experiments on mainstream commercial T2V models including Veo 3.1, Sora 2, Seedance and Kling v1. Results show BSB surpasses all existing jailbreak baselines, delivering an average 18.6% relative gain in attack success rate over the strongest competitor across evaluated models. Our findings identify temporal consistency as an understudied yet vital attack surface for T2V models and verify that structured search facilitates effective vulnerability discovery under constrained query budgets.

Dynamic Spectrum Sharing (DSS) is a cornerstone of next-generation wireless systems, yet existing solutions such as Spectrum Access Systems (SAS) rely on centralized administrators that expose sensitive operational metadata and lack cryptographic transaction accountability. Though SAS administrators, such as Google, have introduced pay-as-you-go pricing models, these approaches still face significant privacy and accountability challenges as DSS evolves toward a more open and large-scale spectrum marketplace. We present SpexPay, a privacy-preserving and auditable pay-as-you-go spectrum usage framework that enforces fine-grained, usage-linked payments without revealing user identities. Spexpay integrates BBS+ verifiable credentials, unlinkable session pseudonyms, and selective-disclosure proofs to enforce privacy-preserving access authorization, while leveraging Solidity-based smart contracts to realize automated and non-repudiable escrow settlement. By recording only pseudonymous usage evidence and hash-chained metering data on-chain, the system achieves strong unlinkability while preserving verifiable accountability and auditability. A full prototype demonstrates low end-to-end latency ($\approx$150 ms) and modest on-chain cost ($\approx$603K gas or $\approx$\$0.9), showing that SpexPay is practical for real-world DSS deployments. We also evaluated the user-side cryptographic operations on a Raspberry Pi 5 to assess scalability and suitability for edge-class hardware. Our code and artifacts are publicly available at https://github.com/iambarat/SpexPay.

Result diversification is crucial for generating informative, non-redundant data summaries and query outputs. Although its various formulations have been extensively studied across an array of data-driven disciplines, existing methods fail to address the privacy concerns that arise when the underlying data is sensitive. In this work, we initiate the study of result diversification under differential privacy, focusing on the max-sum diversification (MSD) problem, a widely adopted model with the objective of maximizing a linear combination of a submodular function, quantifying relevance, and the sum of pairwise distances between selected items, quantifying diversity. We propose differentially private algorithms for MSD under both cardinality and matroid constraints, achieving nearly optimal utility guarantees. At the same time, we design more efficient algorithms that maintain strong guarantees. Notably, the proposed algorithms are faster than existing non-private methods, making them appealing even in non-private settings. Experimental evaluations on real-world datasets demonstrate that the proposed approach achieves utility comparable to that of non-private baselines even under strong privacy guarantees, and significantly improves execution times for cardinality constraints.

Jailbreak attacks on large language models are usually evaluated by attacker-centric metrics such as attack success rate (ASR), yet an attack that breaks a model is not necessarily useful for improving its safety. We propose a defender-centric view of jailbreak evaluation, where attacks are evaluated by the downstream safety improvements they enable when used as red-teaming data for safety training. Building on this view, we introduce A-MESS (Minimal Effective Attack-Subset Selection), a setting-agnostic framework for attributing and selecting jailbreak attacks from black-box subset utility observations. A-MESS estimates AttackSHAP, a Shapley-based score that attributes marginal utility to individual attacks and selects compact attack subsets under user-specified budgets via greedy or surrogate-based optimization. Across controlled utility landscapes and real LLM safety settings, we find that ASR rankings are weakly aligned with defender-centric utility, that AttackSHAP can be estimated accurately with limited utility queries, and that directly optimizing subsets yields stronger safety utility than attacker-centric or attribution-only selection. These results suggest evaluating jailbreak attacks as resources for improving safety, not only as tools for breaking models.

LLM agents can leak privacy (e.g., paths, emails) and credentials (e.g., API keys) as agent observations (e.g., tool outputs, shell logs, and file reads) are appended to provider-bound transcripts. Existing placeholder redaction is brittle: it can miss embedded or cross-turn references, over-redact benign lookalikes, and destroy the structure useful for reasoning. We present SlotGuard, a local transcript boundary that can hide sensitive data while retaining agents' performance. SlotGuard rewrites structural bindings as typed, suffix-aware slots, replaces secrets with format-preserving synthetic values, links cross-turn references with a lightweight session graph, and restores raw values only inside the trusted runtime. On controlled repository-oriented agent transcripts, SlotGuard removes all 20,814 annotated structurally sensitive characters across 9,229 paths and reduces credential leakage to 0.0% across 852 planted values. It remains close to raw-transcript task success across four upstream models, while generic redaction drops to 2.5%. Transcript rewriting takes a median of 14.424~$\mu$s per agent turn. The code is publicly accessible at https://github.com/illinoisdata/SlotGuard.

Digital signatures are fundamental to identity authentication and data integrity in cybersecurity, and the NIST-standardized Digital Signature Algorithm (DSA) frequently appears in the cryptography track of CTF competitions. However, DSA relies on number theory, modular arithmetic, and large-integer computation, making both the algorithm and its associated attacks difficult for beginners to follow. Conventional tools often expose only inputs and outputs, leaving the intermediate computations of signing, verification, and key-recovery attacks opaque. This paper presents a DSA signature analysis and visualisation platform tailored to CTF competitions. The platform provides three main capabilities: basic signature generation and verification, reproduction of common CTF attack methods, and dynamic visualisation of attack workflows. It covers three representative nonce vulnerabilities: nonce reuse, linear nonce leakage, and HNP-based lattice attacks. Stepwise displays and highlighted intermediate values make the underlying computations directly inspectable. Experiments show that the platform correctly reproduces the standard DSA workflow and all three attack scenarios.

It is crucial to safeguard computer networks from evolving network security threats and unknown cyberattacks. An essential tool for protecting computer networks against unknown cyber threats is Network Intrusion Detection System (NIDS). However, NIDS faces a major security concern due to its susceptibility to adversarial attacks. Adversarial attacks aim to deceive NIDS by crafting and injecting adversarial examples into the system. These adversarial inputs can deceive the NIDS into misclassifying benign network traffic as malicious. We developed a resilient hybrid defense mechanism aimed to mitigate the impact of two potent adversarial attacks: Fast Gradient Sign Method (FGSM) and Carlini & Wagner (C&W) attack. Our hybrid defense approach leverages the combined strength of two heuristic defense methods: Adversarial Training (AT) and Gaussian Data Augmentation (GDA). GDA provides multi-directional defense, while AT enhances NIDS robustness against specific adversarial vectors. Under pre-attack scenarios, NIDS demonstrated good accuracy and f1-score. However, in the post-attack scenario, its accuracy significantly dropped under FGSM and C&W attacks (0.2649 and 0.4961, respectively). Our proposed hybrid defense method effectively mitigated these adversarial threats, with post-defense accuracy of 96.57% and 89.20% for FGSM and C&W attacks. We evaluated the defense strategy across a range of epsilon and confidence noise factor values (ranging from 0.0001 to 0.0009). This research provides a good direction for future researchers in the emerging area of adversarial machine learning from a security perspective.

Speech recordings used for dementia detection inherently expose speaker identity, raising critical privacy concerns. Existing methods typically address only singular threats and fail to resolve the privacy–utility trade-off. We propose a multi-level framework designed to neutralize two distinct eavesdropping vectors. At the signal level, a Cumulative Signal Attack (CSA) concentrates perturbations in keyword-aligned regions to maximize transcription error (Word Error Rate WER = 1.00) while preserving vital prosodic biomarkers. At the feature level, a Gradient Reversal Layer (GRL) with Mutual Information (MI)-guided noise injection suppresses speaker-discriminative dimensions while retaining dementia-relevant diagnostic structure. Evaluated on the DementiaBank Pitt Corpus, our framework achieves near-chance speaker identification (Equal Error Rate EER = 0.59, F1 = 0.003) while maintaining strong dementia classification performance (F1 = 0.78, AUC = 0.86).

Routing quantum keys over low-earth-orbit (LEO) satellite constellations is harder than classical routing: satellite handovers couple consecutive scheduling decisions, stochastic cloud cover can silently zero a ground link, and finite-key effects eliminate short, low-elevation passes entirely. We present SATLOCK, a handover-aware Quantum Key Distribution (QKD) routing framework that combines (i) a composite channel model incorporating atmospheric loss, pointing jitter, Markov cloud cover, decoy-state estimation, and finite-key correction; (ii) an integer linear program (ILP) giving a provable handover-aware throughput upper bound; and (iii) a decentralized deep Q-network (DQN) baseline for weather-adaptive online routing. We evaluate two contention regimes on a Walker constellation serving intercontinental demands. In low contention (16 satellites, 6 demands), the ILP delivers 1,311 Mbit while the strongest heuristics reach 95–96% of ILP. In high contention (8 satellites, 12 demands), where handovers become binding, heuristics drop to 89.5% of ILP. The DQN agent reaches 91.8% and 84.6% of ILP in the two regimes; it learns effective per-demand weather policies but is limited in aggregate by the lack of cross-demand coordination.

The advent of LLMs has significantly changed the research on privacy policy and data compliance analysis by enabling tasks that previously required specialized, domain-specific tools. However, it remains unclear to what extent LLMs can truly replicate the diverse functionalities, and the wide range of methodologies and analysis offered by prior work. In this paper, we conduct the first systematic evaluation of whether off-the-shelf LLMs can replace specialized privacy analysis tools. We study six representative tools spanning three major functionalities: contradiction detection, regulatory compliance analysis, and privacy policy summarization and aggregation, and across three intermediate tasks: structured data extraction using tuples, Semantic Role Labeling (SRL) and manual privacy policy labeling. We compare the performance of two state-of-the-art LLMs (GPT-5.2 and Gemini-2.5 in various configurations) against the tools by directly prompting the models to perform corresponding functionalities and tasks on a custom dataset of 10 privacy policies, allowing us to assess whether off-the-shelf models can produce tool-specific functionalities without further engineering or domain-specific training, major limitations in prior work. Our results show that LLMs consistently match or exceed the capabilities of existing tools across the functionalities. In manual labeling of first-party collection entities, LLMs achieved an average precision of 81.8% and recall of 70.9%, while for labeling of third-party sharing entities, they achieved an average precision of 91.4% and recall of 70.8% compared to the OPP-115 dataset. Overall, our findings indicate that LLMs can effectively perform a broad range of functionalities and tasks in privacy policy and regulation analysis that previously required specialized tools.

While the Internet of Things (IoT) has become essential, they introduced serious security and privacy challenges, especially for mission-critical environments. Legacy devices are vulnerable to viruses, data breaches, and unauthorized access, and updating these devices would be infeasibly costly. As a solution, this paper presents a Federated Learning and LLM-Driven Threat Intelligence for Zero Trust IoT Architecture, with FL for anomaly detection integrating privacy-preserving distributed learning, continuous identity verification, and LLM-driven autonomous threat response into a unified pipeline. Unlike existing solutions, our framework enforces Zero Trust at every communication layer via mutual TLS (mTLS) over MQTT, ensuring no device or message is implicitly trusted. Our experiments with Raspberry Pis and various sensors achieve an F1 score of 0.9091 and an ROC-AUC of 1.0, highlighting the effectiveness of the proposed framework in enabling privacy-preserving anomaly detection for resource-constrained IoT devices.

Drone swarms are increasingly deployed in critical applications such as surveillance, disaster response, and infrastructure monitoring. However, their reliance on open communication channels and their limited computational resources make them vulnerable to a wide range of cyber-threats. There is a growing interest in intrusion detection systems (IDS) specifically designed for drone environments and operations. However, the conventional solutions including Machine Learning (ML)-based approaches require collecting all data from heterogeneous drones in the swarm and processing on a central server may not be always feasible. Federated Learning (FL) has emerged as a promising distributed solution with an additional privacy-preserving feature. Even though potential studies exist, conventional FL-based IDS frameworks still face communication and computational overhead challenges, while achieving a balance between efficiency and effective detection under practical resource constraints remains a challenge. Therefore, we propose a lightweight FL-based IDS tailored for drone swarm networks using deep neural networks (DNN) enhanced with knowledge distillation (KD) to reduce model complexity and communication costs without sacrificing detection performance. We evaluate our framework using Raspberry Pi 4 devices and a real-world drone network dataset. Our approach demonstrates a detection accuracy of approximately 98.6% while reducing overall communication cost by around 70% and computational overhead by 29%. These results show that FL combined with KD is a practical and suitable solution for secure and efficient deployment in resource-constrained drone networks.

Large security operations centers (SOCs) often face hundreds of active incidents per day, creating substantial cognitive and operational demands for analysts. Analysts must quickly decide which incidents deserve attention within long, constantly changing queues, yet incidents are commonly ordered by arrival time, coarse severity, or product-specific heuristics that leave their relative priority unclear. We introduce Adaptive Incident Prioritization (AIP), the ranking algorithm behind Microsoft Defender Queue Assistant, which continuously prioritizes security incidents for analyst investigation. AIP adapts BM25-style ranking to a query-less, multi-tenant queue setting by representing each incident as a collection of normalized security components extracted from alerts and metadata. The model combines saturated local component frequency, global component rarity estimated across tenants, bounded domain-prior multipliers, and component-level explanations. Deployed across tens of thousands of customers, AIP performs near-real-time inference and refreshes incident scores with a median latency of five seconds. In an expert-reviewed evaluation across 1,000 customer organizations, AIP achieves 92.8% Precision@10. In post-launch telemetry across 473,000 organization-day queues, AIP increases alert-detail interaction by 5.8% and alert-detail view events by 17.5% relative to severity ordering, providing behavioral evidence that model-ranked queues concentrate analyst engagement. We also extend the Microsoft GUIDE dataset with, to our knowledge, the first public label source for SOC queue prioritization over real-world incidents. The extension covers 499 organization queues and 9,980 incidents with expert-derived priority labels, enabling the research community to develop, compare, and advance methods for incident prioritization.

Blockchain platforms have grown into an ecosystem of independent networks, and a growing class of applications now requires smart contracts on separate chains to act as one. Such operations must be atomic, yet immutability makes this fundamentally harder: a confirmed transaction cannot be reversed, so the rollback on which classical atomic commitment protocols depend is unavailable. Two challenges follow. Contract state must be held across an operation whose outcome is not yet known, and each chain's execution outcome must be established even though no chain can observe another. In response, we introduce a framework that achieves atomicity through forward-only correction, resolving incomplete operations with new on-chain transactions rather than reversal. The framework bounds how long contract state is held and confines contention to the state an operation touches, and it establishes outcomes from an on-chain record of what each chain executed, without relying on any single coordinating party. This work lays the foundation for atomic coordination of general smart contract operations across heterogeneous blockchains.

Unified multimodal models (UMMs) have recently demonstrated powerful instruction-based image editing capabilities, while also raising serious concerns about the unauthorized manipulation of personal portraits. We investigate a novel and practical problem: protecting facial identities against unauthorized editing of UMMs. Existing diffusion-based and VLM-based protection methods often become ineffective because they typically disrupt only a single visual branch. To understand this limitation, we conduct a feature-level analysis of the understanding and generation branches in unified image editing models. Our observations show that the structural agreement between these two branches is closely related to successful image editing. When only one branch is distorted, the model may still recover identity information from the other branch. Based on this, we propose Cross-Branch Conflict as a Shield (CCS), a unified adversarial protection framework. CCS jointly drives the ViT and VAE representations away from their clean counterparts. It also uses a linear Centered Kernel Alignment (CKA) objective to disrupt the structural consistency between the two branches. By degrading reliable identity information in both visual pathways and inducing incompatible cross-branch representations, CCS effectively prevents UMMs from recovering consistent facial identity cues during editing. Extensive experiments suggest that CCS consistently provides stronger protection in suppressing identity-preserving edits. Codes are in the supplementary material.

End-to-end speech language models increasingly represent user speech with speech tokens rather than relying exclusively on cascaded ASR–LLM–TTS pipelines. Although these tokens support expressive and low-latency spoken interaction, they may also preserve sensitive speaker characteristics. We investigate whether exposed speech tokens leak voiceprints and formulate this risk as a speaker inversion attack. We introduce Audio BERT (AuB), a trainable model that constructs token embeddings from discrete codebooks and aggregates them into speaker-sensitive representations, and propose SpInv, a two-stage inversion method built on AuB to recover embeddings in the space of an attacker-specified speaker encoder. We evaluate Moshi, Higgs3, Kimi-Audio, and Qwen3-Omni using speaker-disjoint protocols on the VoxCeleb dataset. Extensive experiments show that, with only three seconds of frontend output, SpInv achieves cosine similarities above 0.70 in the attacker-specified speaker-encoder space.

Conventional processor designs expose code and data as plaintext throughout execution, rendering them inherently vulnerable to attacks that recover intellectual property or modify security/safety checks. Instruction-level encryption (ILE) enables CPU-level decryption, execution, and optionally authentication of individual encrypted program instructions at runtime. However, existing proposals depend on specific micro-architectures, detect corrupted instructions after they have executed, rely on non-standard ciphers, or require complex analyses of program state. In this work, we introduce and present a design exploration of a RISC-V processor architecture (SABLE) that enables minimally invasive instruction-level authenticated encryption of programs. SABLE is agnostic to the underlying micro-architecture, remaining compatible with the standard RISC-V toolchain with minor changes to post-process compiled ELF binaries. We integrate a decrypt-and-verify stage at two points (the instruction-memory wrapper and the CPU frontend) and explore seven ILE micro-architectures from a single-cycle (combinational) design to six multi-cycle (sequential) variants. We implement and evaluate the designs using ASCON-128a on a Xilinx Artix-7 FPGA with the open-source NEORV32 system on chip. Relative to baseline performance, the configurations span LUT, performance, power, and energy-per-instruction overheads of 1.6-9.3$\times$, 4.1-10.0$\times$, 1.5-8.0$\times$, and 10.4-80.0$\times$, respectively, using the Dhrystone benchmarking suite. Finally, we discuss design trade-offs, highlighting area-, performance-, and energy-aware design points.

Authorization violations under valid Web sessions are difficult to identify and handle in real time from traffic because they depend strongly on business semantics and exhibit few distinctive protocol-level features. This paper proposes a non-intrusive traffic analysis framework for authorization risk detection and coordinated response. Request-response transactions are correlated to extract the access subject, business endpoint, object identifier, authentication state, and behavioral sequence. Object-access evidence, identity consistency, behavioral anomalies, authentication context, network environment, and endpoint-operation risk are mapped to interpretable risk components. Weighted fusion and high-risk priority constraints produce graded decisions that drive allow, alert, block, and external policy actions. The risk components are instantiated using deterministic and interpretable rules to evaluate multi-source evidence organization, risk fusion, and the coordinated-response loop at the framework level. In a controlled local testbed containing 2,000 balanced labeled samples, the framework classified 998 of 1,000 authorization-risk events as risky and produced no false positives among 1,000 normal accesses, achieving 99.90% accuracy, 100.00% precision, 99.80% recall, and a 99.90% F1 score. Removing runtime object evidence reduced the F1 score to 81.31%, while removing the high-risk priority constraint reduced it to 73.90%. In a prototype feasibility test with 100 concurrent requests and 1,000 total requests, mean risk-decision computation latency was 0.077 ms and P99 latency was 0.137 ms. The results demonstrate the feasibility of organizing heterogeneous runtime evidence and establishing an executable authorization-risk decision and coordinated-response loop without modifying application code, but do not establish general applicability in production environments.

The increasing adoption of Large Language Models (LLMs) as AI components in modern software systems introduces distinct security risks to the software supply chain. While many considerations and safety mechanisms are in place for components of the traditional software supply chain, the recent rapid adoption of AI components and platforms has overlooked these hard learned lessons. Selecting and integrating AI models without clear guidance on how these choices affect system security may leave applications vulnerable to threats, such as malicious components, data leakage, and unintended behavior. The goal of this study is to understand practitioners' decision making process and security considerations in selecting and integrating AI components through an exploratory semi-structured interview study. Toward this goal, we conducted semistructured interviews with 22 software developers, architects, and AI practitioners across diverse organizations about how they integrate AI components into their software. Our analysis finds that practitioners' model selection is predominantly driven by functional criteria, including performance, accuracy, cost, and specific features, e.g., tool calling or multimodal support, while security is rarely considered as an evaluation criterion. We observe a consistent lack of security concern throughout the AI component integration process, with established software supply chain lessons overlooked or ignored. The industry is repeating the historically costly mistakes of early software dependency management, prioritizing rapid reuse and availability over security and provenance. We distill our findings into actionable recommendations for AI adopters, model providers, and researchers, advocating for a proactive, security-by-design approach that integrates security evaluation into component selection and sustains it throughout the software development lifecycle.