Papers

Quantum Key Distribution (QKD) is revolutionizing cryptography by promising information-theoretic security through the immutable laws of quantum mechanics. Yet, the challenge of transforming these idealized security models into practical, resilient systems remains a pressing issue, especially as quantum computing evolves. In this review, we critically dissect and synthesize the latest advancements in QKD protocols and their security vulnerabilities, with a strong emphasis on rigorous security proofs. We actively categorize contemporary QKD schemes into three key classes: uncertainty principle-based protocols (e.g., BB84), hybrid architectures that enable secure direct communication (eg, three-stage protocol), and continuous-variable frameworks. We further include two modern classes of QKD protocols, namely Twin-field QKD and Device-Independent QKD, both of which were developed to have practical implementations over the last decade. Moreover, we highlight important experimental breakthroughs and innovative mitigation strategies, including the deployment of advanced Quantum Error Correction Codes (QECCs), that significantly enhance channel fidelity and system robustness. By mapping the current landscape, from sophisticated quantum attacks to state-of-the-art error correction methods, this review fills an important gap in the literature. To bring everything together, the relevance of this review concerning quantum augmented networks (QuANets) is also presented. This allows the readers to gain a comprehensive understanding of the security promises of quantum key distribution from theoretical proofs to experimental validations.

The integration of artificial intelligence (AI) agents into web browsers introduces security challenges that go beyond traditional web application threat models. Prior work has identified prompt injection as a new attack vector for web agents, yet the resulting impact within real-world environments remains insufficiently understood. In this work, we examine the landscape of prompt injection attacks and synthesize a benchmark of attacks embedded in realistic HTML payloads. Our benchmark goes beyond prior work by emphasizing injections that can influence real-world actions rather than mere text outputs, and by presenting attack payloads with complexity and distractor frequency similar to what real-world agents encounter. We leverage this benchmark to conduct a comprehensive empirical evaluation of existing defenses, assessing their effectiveness across a suite of frontier AI models. We propose a multi-layered defense strategy comprising both architectural and model-based defenses to protect against evolving prompt injection attacks. Our work offers a blueprint for designing practical, secure web agents through a defense-in-depth approach.

Command-line interface (CLI) fuzzing tests programs by mutating both command-line options and input file contents, thus enabling discovery of vulnerabilities that only manifest under specific option-input combinations. Prior works of CLI fuzzing face the challenges of generating semantics-rich option strings and input files, which cannot reach deeply embedded target functions. This often leads to a misdetection of such a deep vulnerability using existing CLI fuzzing techniques. In this paper, we design a novel Path-guided, Iterative LLM-Orchestrated Testing framework, called PILOT, to fuzz CLI applications. The key insight is to provide potential call paths to target functions as context to LLM so that it can better generate CLI option strings and input files. Then, PILOT iteratively repeats the process, and provides reached functions as additional context so that target functions are reached. Our evaluation on real-world CLI applications demonstrates that PILOT achieves higher coverage than state-of-the-art fuzzing approaches and discovers 51 zero-day vulnerabilities. We responsibly disclosed all the vulnerabilities to their developers and so far 41 have been confirmed by their developers with 33 being fixed and three assigned CVE identifiers.

Securing the Internet of Things (IoT) against quantum attacks requires public-key cryptography that (i) remains compact and (ii) runs efficiently on microcontrollers, capabilities many post-quantum (PQ) schemes lack due to large keys and heavy arithmetic. We address both constraints simultaneously with, to our knowledge, the first-ever isogeny framework that encodes super-singular elliptic-curve isogeny data via novel Engel expansions over the p-adic Laurent series. Engel coefficients compress torsion information, thereby addressing the compactness constraint, yielding public keys of ~1.1 - 16.9 kbits preserving the hallmark small sizes of isogeny systems. Engel arithmetic is local and admits fixed-precision p-adic operations, enabling micro-controller efficiency with low-memory, branch-regular kernels suitable for embedded targets.

Autonomous digital entities require deterministic identity mechanisms that avoid persistent storage of high-value master secrets, while supporting credential rotation and cryptographic agility across heterogeneous systems. Existing deterministic key hierarchies and centralized key management systems typically rely on long-lived root secrets, introducing structural single points of failure and complicating lifecycle management. We present ACE-GF (Atomic Cryptographic Entity Generative Framework), a seed-storage-free identity construction that enables deterministic and context-isolated key derivation without storing any master secret at rest. The construction reconstructs an identity root ephemerally in memory from a sealed artifact and authorization credentials, using misuse-resistant authenticated encryption together with standard key derivation primitives. Derived keys are generated via HKDF with explicit context encoding, ensuring cryptographic isolation across curves and application domains. This design naturally supports stateless credential rotation, authorization-bound revocation, and non-disruptive migration toward post-quantum cryptographic domains. Furthermore, the framework's parametric agility allows for optimization in resource-constrained environments, ensuring that deterministic identity reconstruction remains viable across a spectrum of hardware from high-performance servers to low-power IoT nodes without compromising the underlying security model. This work builds upon the conceptual framework introduced in MSCIKDF, which identified the core design goals for multi-curve, context-isolated, PQC-pluggable identity but did not provide a concrete construction. A formal protocol specification of ACE-GF has been submitted as an IETF Internet-Draft.

Advanced Persistent Threats (APT) pose a major cybersecurity challenge due to their stealth, persistence, and adaptability. Traditional machine learning detectors struggle with class imbalance, high dimensional features, and scarce real world traces. They often lack transferability-performing well in the training domain but degrading in novel attack scenarios. We propose a hybrid transfer framework that integrates Transfer Learning, Explainable AI (XAI), contrastive learning, and Siamese networks to improve cross-domain generalization. An attention-based autoencoder supports knowledge transfer across domains, while Shapley Additive exPlanations (SHAP) select stable, informative features to reduce dimensionality and computational cost. A Siamese encoder trained with a contrastive objective aligns source and target representations, increasing anomaly separability and mitigating feature drift. We evaluate on real-world traces from the DARPA Transparent Computing (TC) program and augment with synthetic attack scenarios to test robustness. Across source to target transfers, the approach delivers improved detection scores with classical and deep baselines, demonstrating a scalable, explainable, and transferable solution for APT detection.

Advanced Persistent Threats (APTs) pose a significant challenge in cybersecurity due to their stealthy and long-term nature. Modern supervised learning methods require extensive labeled data, which is often scarce in real-world cybersecurity environments. In this paper, we propose an innovative approach that leverages AutoEncoders for unsupervised anomaly detection, augmented by active learning to iteratively improve the detection of APT anomalies. By selectively querying an oracle for labels on uncertain or ambiguous samples, we minimize labeling costs while improving detection rates, enabling the model to improve its detection accuracy with minimal data while reducing the need for extensive manual labeling. We provide a detailed formulation of the proposed Attention Adversarial Dual AutoEncoder-based anomaly detection framework and show how the active learning loop iteratively enhances the model. The framework is evaluated on real-world imbalanced provenance trace databases produced by the DARPA Transparent Computing program, where APT-like attacks constitute as little as 0.004% of the data. The datasets span multiple operating systems, including Android, Linux, BSD, and Windows, and cover two attack scenarios. The results have shown significant improvements in detection rates during active learning and better performance compared to other existing approaches.

The Software Bill of Materials (SBOM) is a critical tool for securing the software supply chain (SSC), but its practical utility is undermined by inaccuracies in both its generation and its application in vulnerability scanning. This paper presents a large-scale empirical study on 2,414 open-source repositories to address these issues from a practical standpoint. First, we demonstrate that using lock files with strong package managers enables the generation of accurate and consistent SBOMs, establishing a reliable foundation for security analysis. Using this high-fidelity foundation, however, we expose a more fundamental flaw in practice: downstream vulnerability scanners produce a staggering 92.0% false positive rate in our case study. We pinpoint the primary cause as the flagging of vulnerabilities within unreachable code. We then demonstrate that function call analysis can effectively prune 61.9% of these false alarms. Our work validates a practical, two-stage approach for SSC security: first, generate an accurate SBOM using lock files and strong package managers, and second, enrich it with function call analysis to produce actionable, low-noise vulnerability reports that alleviate developers' alert fatigue.

Provenance-based threat hunting identifies Advanced Persistent Threats (APTs) on endpoints by correlating attack patterns described in Cyber Threat Intelligence (CTI) with provenance graphs derived from system audit logs. A fundamental challenge in this paradigm lies in the modality gap – the structural and semantic disconnect between provenance graphs and CTI reports. Prior work addresses this by framing threat hunting as a graph matching task: 1) extracting attack graphs from CTI reports, and 2) aligning them with provenance graphs. However, this pipeline incurs severe information loss during graph extraction and demands intensive manual curation, undermining scalability and effectiveness. In this paper, we present APT-CGLP, a novel cross-modal APT hunting system via Contrastive Graph-Language Pre-training, facilitating end-to-end semantic matching between provenance graphs and CTI reports without human intervention. First, empowered by the Large Language Model (LLM), APT-CGLP mitigates data scarcity by synthesizing high-fidelity provenance graph-CTI report pairs, while simultaneously distilling actionable insights from noisy web-sourced CTIs to improve their operational utility. Second, APT-CGLP incorporates a tailored multi-objective training algorithm that synergizes contrastive learning with inter-modal masked modeling, promoting cross-modal attack semantic alignment at both coarse- and fine-grained levels. Extensive experiments on four real-world APT datasets demonstrate that APT-CGLP consistently outperforms state-of-the-art threat hunting baselines in terms of accuracy and efficiency.

Precise access control decisions are crucial for the security of both traditional applications and emerging agent-based systems. Typically, these decisions are made by users during app installation or at runtime. However, due to the increasing complexity and automation of systems, making access control decisions can impose a significant cognitive burden on users, often overwhelming them and leading to suboptimal or even arbitrary choices. To address this problem, we investigate the ability of LLMs to make dynamic, context-aware decisions aligned with users' security preferences, expressed during a lightweight setup phase. As a case study, we analyze smartphone application permission requests, given their ubiquity and users' familiarity with them. We curated a dataset comprising 307 user privacy statements (short, natural-language descriptions of user preferences) and 14,682 corresponding permission decisions, gathered from smartphone users in an online data collection. We compare these decisions with those made by two versions of LLMs that are tasked with reasoning about the app and the request context: a general model and a personalized one (which incorporates user preferences). For the latter, we also collected user feedback on 1,298 of its decisions. Our results show that LLMs generally reflect users' preferences well, agreeing with the majority decision in up to 86% of cases, and can steer users toward safer behavior. However, the results also reveal a key trade-off in personalization: while incorporating user-specific privacy preferences improves agreement with individual decisions, strict adherence to these preferences may lead to less safe outcomes, as users tend to over-permission.

WhatsApp, with 3.5 billion active accounts as of early 2025, is the world's largest instant messaging platform. Given its massive user base, WhatsApp plays a critical role in global communication. To initiate conversations, users must first discover whether their contacts are registered on the platform. This is achieved by querying WhatsApp's servers with mobile phone numbers extracted from the user's address book (if they allowed access). This architecture inherently enables phone number enumeration, as the service must allow legitimate users to query contact availability. While rate limiting is a standard defense against abuse, we revisit the problem and show that WhatsApp remains highly vulnerable to enumeration at scale. In our study, we were able to probe over a hundred million phone numbers per hour without encountering blocking or effective rate limiting. Our findings demonstrate not only the persistence but the severity of this vulnerability. We further show that nearly half of the phone numbers disclosed in the 2021 Facebook data leak are still active on WhatsApp, underlining the enduring risks associated with such exposures. Moreover, we were able to perform a census of WhatsApp users, providing a glimpse on the macroscopic insights a large messaging service is able to generate even though the messages themselves are end-to-end encrypted. Using the gathered data, we also discovered the re-use of certain X25519 keys across different devices and phone numbers, indicating either insecure (custom) implementations, or fraudulent activity. In this updated version of the paper, we also provide insights into the collaborative remediation process through which we confirmed that the underlying rate-limiting issue had been resolved.

Nowadays, malware increasingly uses DNS-based covert channels in order to evade detection and maintain stealthy communication with its command-and-control servers. While prior work has focused on detecting such activity, identifying specific malware families and their behaviors from captured network traffic remains challenging due to the variability of DNS. In this paper, we present the first application of Locality Sensitive Hashing to the detection and identification of real-world malware utilizing DNS covert channels. Our approach encodes DNS subdomain sequences into statistical similarity features that effectively capture anomalies indicative of malicious activity. Combined with a Random Forest classifier, our method achieves higher accuracy and reduced false positive rates than prior approaches, while demonstrating improved robustness and generalization to previously unseen or modified malware samples. We further demonstrate that our approach enables reliable classification of malware behavior (e.g., uploading or downloading of files), based solely on DNS subdomains.

Multimodal Large Language Models (MLLMs) are increasingly deployed in stateless systems, such as autonomous driving and robotics. This paper investigates a novel threat: Semantic-Aware Hijacking. We explore the feasibility of hijacking multiple stateless decisions simultaneously using a single universal perturbation. We introduce the Semantic-Aware Universal Perturbation (SAUP), which acts as a semantic router, "actively" perceiving input semantics and routing them to distinct, attacker-defined targets. To achieve this, we conduct theoretical and empirical analysis on the geometric properties in the latent space. Guided by these insights, we propose the Semantic-Oriented (SORT) optimization strategy and annotate a new dataset with fine-grained semantics to evaluate performance. Extensive experiments on three representative MLLMs demonstrate the fundamental feasibility of this attack, achieving a 66% attack success rate over five targets using a single frame against Qwen.

Recent advances in artificial intelligence (AI), particularly deep learning, have led to widespread adoption across various applications. Yet, a fundamental challenge persists: how can we verify the correctness of AI model inference when model owners cannot (or will not) reveal their parameters? These parameters represent enormous training costs and valuable intellectual property, making transparent verification difficult. In this paper, we introduce a zero-knowledge framework capable of verifying deep learning inference without exposing model internal parameters. Built on recursively composed zero-knowledge proofs and requiring no trusted setup, our framework supports both linear and nonlinear neural network layers, including matrix multiplication, normalization, softmax, and SiLU. Leveraging the Fiat-Shamir heuristic, we obtain a succinct non-interactive argument of knowledge (zkSNARK) with constant-size proofs. To demonstrate the practicality of our approach, we translate the DeepSeek model into a fully SNARK-verifiable version named ZK-DeepSeek and show experimentally that our framework delivers both efficiency and flexibility in real-world AI verification workloads.

As deep image forgery powered by AI generative models, such as GANs, continues to challenge today's digital world, detecting AI-generated forgeries has become a vital security topic. Generalizability and robustness are two critical concerns of a forgery detector, determining its reliability when facing unknown GANs and noisy samples in an open world. Although many studies focus on improving these two properties, the root causes of these problems have not been fully explored, and it is unclear if there is a connection between them. Moreover, despite recent achievements in addressing these issues from image forensic or anti-forensic aspects, a universal method that can contribute to both sides simultaneously remains practically significant yet unavailable. In this paper, we provide a fundamental explanation of these problems from a frequency perspective. Our analysis reveals that the frequency bias of a DNN forgery detector is a possible cause of generalization and robustness issues. Based on this finding, we propose a two-step frequency alignment method to remove the frequency discrepancy between real and fake images, offering double-sided benefits: it can serve as a strong black-box attack against forgery detectors in the anti-forensic context or, conversely, as a universal defense to improve detector reliability in the forensic context. We also develop corresponding attack and defense implementations and demonstrate their effectiveness, as well as the effect of the frequency alignment method, in various experimental settings involving twelve detectors, eight forgery models, and five metrics.

As AI agents become integral to enterprise workflows, their reliance on shared tool libraries and pre-trained components creates significant supply chain vulnerabilities. While previous work has demonstrated behavioral backdoor detection within individual LLM architectures, the critical question of cross-LLM generalization remains unexplored, a gap with serious implications for organizations deploying multiple AI systems. We present the first systematic study of cross-LLM behavioral backdoor detection, evaluating generalization across six production LLMs (GPT-5.1, Claude Sonnet 4.5, Grok 4.1, Llama 4 Maverick, GPT-OSS 120B, and DeepSeek Chat V3.1). Through 1,198 execution traces and 36 cross-model experiments, we quantify a critical finding: single-model detectors achieve 92.7% accuracy within their training distribution but only 49.2% across different LLMs, a 43.4 percentage point generalization gap equivalent to random guessing. Our analysis reveals that this gap stems from model-specific behavioral signatures, particularly in temporal features (coefficient of variation > 0.8), while structural features remain stable across architectures. We show that model-aware detection incorporating model identity as an additional feature achieves 90.6% accuracy universally across all evaluated models. We release our multi-LLM trace dataset and detection framework to enable reproducible research.

Large language model (LLM) services have been rapidly integrated into people's daily lives as chatbots and agentic systems. They are nourished by collecting rich streams of data, raising privacy concerns around excessive collection of sensitive personal information. Privacy policies are the fundamental mechanism for informing users about data practices in modern information privacy paradigm. Although traditional web and mobile policies are well studied, the privacy policies of LLM providers, their LLM-specific content, and their evolution over time remain largely underexplored. In this paper, we present the first longitudinal empirical study of privacy policies for mainstream LLM providers worldwide. We curate a chronological dataset of 74 historical privacy policies and 115 supplemental privacy documents from 11 LLM providers across 5 countries up to August 2025, and extract over 3,000 sentence-level edits between consecutive policy versions. We compare LLM privacy policies to those of other software formats, propose a taxonomy tailored to LLM privacy policies, annotate policy edits and align them with a timeline of key LLM ecosystem events. Results show they are substantially longer, demand college-level reading ability, and remain highly vague. Our taxonomy analysis reveals patterns in how providers disclose LLM-specific practices and highlights regional disparities in coverage. Policy edits are concentrated in first-party data collection and international/specific-audience sections, and that product releases and regulatory actions are the primary drivers, shedding light on the status quo and the evolution of LLM privacy policies.

The integration of Large Language Models (LLMs) into healthcare demands a safety paradigm rooted in primum non nocere. However, current alignment techniques rely on generic definitions of harm that fail to capture context-dependent violations, such as administrative fraud and clinical discrimination. To address this, we introduce Medical Malice: a dataset of 214,219 adversarial prompts calibrated to the regulatory and ethical complexities of the Brazilian Unified Health System (SUS). Crucially, the dataset includes the reasoning behind each violation, enabling models to internalize ethical boundaries rather than merely memorizing a fixed set of refusals. Using an unaligned agent (Grok-4) within a persona-driven pipeline, we synthesized high-fidelity threats across seven taxonomies, ranging from procurement manipulation and queue-jumping to obstetric violence. We discuss the ethical design of releasing these "vulnerability signatures" to correct the information asymmetry between malicious actors and AI developers. Ultimately, this work advocates for a shift from universal to context-aware safety, providing the necessary resources to immunize healthcare AI against the nuanced, systemic threats inherent to high-stakes medical environments – vulnerabilities that represent the paramount risk to patient safety and the successful integration of AI in healthcare systems.

In the age of agentic AI, the growing deployment of multi-modal models (MMs) has introduced new attack vectors that can leak sensitive training data in MMs, causing privacy leakage. This paper investigates a black-box privacy attack, i.e., membership inference attack (MIA) on multi-modal vision-language models (VLMs). State-of-the-art research analyzes privacy attacks primarily to unimodal AI-ML systems, while recent studies indicate MMs can also be vulnerable to privacy attacks. While researchers have demonstrated that biologically inspired neural network representations can improve unimodal model resilience against adversarial attacks, it remains unexplored whether neuro-inspired MMs are resilient against privacy attacks. In this work, we introduce a systematic neuroscience-inspired topological regularization (tau) framework to analyze MM VLMs resilience against image-text-based inference privacy attacks. We examine this phenomenon using three VLMs: BLIP, PaliGemma 2, and ViT-GPT2, across three benchmark datasets: COCO, CC3M, and NoCaps. Our experiments compare the resilience of baseline and neuro VLMs (with topological regularization), where the tau > 0 configuration defines the NEURO variant of VLM. Our results on the BLIP model using the COCO dataset illustrate that MIA attack success in NEURO VLMs drops by 24% mean ROC-AUC, while achieving similar model utility (similarities between generated and reference captions) in terms of MPNet and ROUGE-2 metrics. This shows neuro VLMs are comparatively more resilient against privacy attacks, while not significantly compromising model utility. Our extensive evaluation with PaliGemma 2 and ViT-GPT2 models, on two additional datasets: CC3M and NoCaps, further validates the consistency of the findings. This work contributes to the growing understanding of privacy risks in MMs and provides evidence on neuro VLMs privacy threat resilience.

Large language models (LLMs) and LLM-based coding agents are now used to generate code from natural-language specifications, yet ensuring such code is both functionally correct and secure remains a challenge. We present DualGauge, the first fully automated framework for jointly evaluating correctness and security of specification-only code generation, supported by DualGauge-Bench, a language-agnostic benchmark of 307 coding tasks each paired with functional and security tests derived from the same specification. Evaluating 10 representative LLMs across Python, C++, and JavaScript, we find that functional correctness substantially overestimates reliable code generation: even the strongest model remains below 15% joint security-functionality success in every language. Common model-side factors–scale, extended thinking, quantization, instruction tuning, and code specialization–do not reliably improve joint performance, suggesting secure-and-correct code generation does not simply emerge from stronger coding capability. Evaluation of 3 leading coding agents (Codex, OpenHands, and Claude Code) shows that iterative scaffolding is not a reliable remedy on specification-only tasks: results are strongly scaffold-dependent, with two of the three agents failing to outperform a comparable direct model. A qualitative audit reveals that these failures concentrate at the output contract boundary and in guards that exist but are insufficient–patterns that only joint benchmarking reliably exposes.

Large Language Models (LLMs) remain vulnerable to prompt injection attacks, representing the most significant security threat in production deployments. We present Prompt Fencing, a novel architectural approach that applies cryptographic authentication and data architecture principles to establish explicit security boundaries within LLM prompts. Our approach decorates prompt segments with cryptographically signed metadata including trust ratings and content types, enabling LLMs to distinguish between trusted instructions and untrusted content. While current LLMs lack native fence awareness, we demonstrate that simulated awareness through prompt instructions achieved complete prevention of injection attacks in our experiments, reducing success rates from 86.7% (260/300 successful attacks) to 0% (0/300 successful attacks) across 300 test cases with two leading LLM providers. We implement a proof-of-concept fence generation and verification pipeline with a total overhead of 0.224 seconds (0.130s for fence generation, 0.094s for validation) across 100 samples. Our approach is platform-agnostic and can be incrementally deployed as a security layer above existing LLM infrastructure, with the expectation that future models will be trained with native fence awareness for optimal security.

MPC-based ML uses multi-party computation (MPC) to run machine learning (ML) workloads across multiple parties without each having to share their private data or model parameters. However, existing frameworks frequently degrade accuracy and performance due to a series of MPC-specific transformations that add errors and overheads. These transformations are mostly opaque to users, making it hard to find and/or optimize problematic transformations. We propose CrypTorch, a modular, extensible, and iteratively-testable compiler framework for MPC-based ML. CrypTorch splits MPC-specific transformations into modular compilation stages, allowing users to easily inspect and optimize them. CrypTorch emits an executable graph after each transformation, allowing iterative testing to pinpoint any problematic transformations. Building on these features, CrypTorch automatically chooses a set of transformations from a pool of choices to balance performance and accuracy during the operator approximation stage, which we identified as the biggest contributor to accuracy/performance degradation. CrypTorch's auto-tuning alone provides 1.21-1.5x speedup without accuracy loss, and 1.33-1.74x speedup when some accuracy degradation is allowed. Combined with better engineering and adoption of state-of-the-art practices (made easier due to CrypTorch's modular design), CrypTorch brings 3.74-8.32x end-to-end speedup compared to the popular CrypTen. CrypTorch is built as an extension to PyTorch 2's compiler.

Linux-based systems rely on large ecosystems of packaged binaries, many of which are implemented in C/C++. While these languages provide low-level memory control and high performance, they are also highly susceptible to memory-safety vulnerabilities, particularly buffer overflows (BOs). Traditional vulnerability discovery techniques, such as static and dynamic analysis, often struggle with scalability and precision when applied directly to binary code, which can thereby keep programs vulnerable. This work introduces BASICS, an approach and framework that combines model checking and bounded concolic execution to automatically verify security properties of stack memory in binary code, trampoline-based binary repair and crash-inducing inputs for bounded patch validation. BASICS constructs a Memory State Space – MemStaCe – from the binary program's control-flow graph and bounded concolic execution of C function calls and loops. Security properties, defined in Linear Temporal Logic, model vulnerable stack behaviours and allow the approach to identify violations in MemStaCe through counterexample traces. These vulnerabilities are then addressed using trampoline-based binary patching, and the resulting patches are validated using crash-inducing inputs extracted during concolic execution. We evaluated BASICS on the Juliet C/C++ and SARD benchmarks, as well as real open-source and Linux package binaries. BASICS achieved 86.4% precision and 87.4% accuracy on SARD, and 86.1% precision on Juliet, while also generating validated binary patches for supported vulnerable sinks. Compared with other evaluated tools, it provided the strongest overall balance between precision, analysis cost, and patching capability.

This study examines whether Low-Rank Adaptation (LoRA) fine-tuned Large Language Models (LLMs) can approximate the performance of fully fine-tuned models in generating human-interpretable decisions and explanations for malware classification. Achieving trustworthy malware detection, particularly when LLMs are involved, remains a significant challenge. We developed an evaluation framework using Bilingual Evaluation Understudy (BLEU), Recall-Oriented Understudy for Gisting Evaluation (ROUGE), and Semantic Similarity Metrics to benchmark explanation quality across five LoRA configurations and a fully fine-tuned baseline. Results indicate that full fine-tuning achieves the highest overall scores, with BLEU and ROUGE improvements of up to 10% over LoRA variants. However, mid-range LoRA models deliver competitive performance exceeding full fine-tuning on two metrics while reducing model size by approximately 81% and training time by over 80% on a LoRA model with 15.5% trainable parameters. These findings demonstrate that LoRA offers a practical balance of interpretability and resource efficiency, enabling deployment in resource-constrained environments without sacrificing explanation quality. By providing feature-driven natural language explanations for malware classifications, this approach enhances transparency, analyst confidence, and operational scalability in malware detection systems.

The ever-increasing number of Android devices and the accelerated evolution of malware, reaching over 35 million samples by 2024, highlight the critical importance of effective detection methods. Attackers are now using Artificial Intelligence to create sophisticated malware variations that can easily evade traditional detection techniques. Although machine learning has shown promise in malware classification, its success relies heavily on the availability of up-to-date, high-quality datasets. The scarcity and high cost of obtaining and labeling real malware samples presents significant challenges in developing robust detection models. In this paper, we propose MalSynGen, a Malware Synthetic Data Generation methodology that uses a conditional Generative Adversarial Network (cGAN) to generate synthetic tabular data. This data preserves the statistical properties of real-world data and improves the performance of Android malware classifiers. We evaluated the effectiveness of this approach using various datasets and metrics that assess the fidelity of the generated data, its utility in classification, and the computational efficiency of the process. Our experiments demonstrate that MalSynGen can generalize across different datasets, providing a viable solution to address the issues of obsolescence and low quality data in malware detection.

Modern enterprise systems face escalating cyber threats that are increasingly dynamic, distributed, and multi-stage in nature. Traditional intrusion detection and response systems often rely on static rules and manual workflows, which limit their ability to respond with the speed and precision required in high-stakes environments. To address these challenges, we present the Intrusion Response System Digital Assistant (IRSDA), an agent-based framework designed to deliver autonomous and policy-compliant cyber defense. IRSDA combines Self-Adaptive Autonomic Computing Systems (SA-ACS) with the Knowledge guided Monitor, Analyze, Plan, and Execute (MAPE-K) loop to support real-time, partition-aware decision-making across enterprise infrastructure. IRSDA incorporates a knowledge-driven architecture that integrates contextual information with AI-based reasoning to support system-guided intrusion response. The framework leverages retrieval mechanisms and structured representations to inform decision-making while maintaining alignment with operational policies. We assess the system using a representative real-world microservices application, demonstrating its ability to automate containment, enforce compliance, and provide traceable outputs for security analyst interpretation. This work outlines a modular and agent-driven approach to cyber defense that emphasizes explainability, system-state awareness, and operational control in intrusion response.

Text-to-image diffusion models can emit copyrighted, unsafe, or private content. Safety alignment aims to suppress specific concepts, yet evaluations seldom test whether safety persists under benign downstream fine-tuning routinely applied after deployment (e.g., LoRA personalization, style/domain adapters). We study the stability of current safety methods under benign fine-tuning and observe frequent breakdowns. As true safety alignment must withstand even benign post-deployment adaptations, we introduce the SPQR benchmark (Safety, Prompt adherence, Quality, and Robustness). SPQR is a single-scored metric that provides a unified, reproducible framework to evaluate how well safety-aligned diffusion models preserve safety, utility, and robustness under benign fine-tuning, by reporting a single leaderboard score to facilitate comparisons. We conduct multilingual, domain-specific, and out-of-distribution analyses, along with category-wise breakdowns, to identify when safety alignment fails after benign fine-tuning, ultimately showcasing SPQR as a concise yet comprehensive benchmark for T2I safety alignment techniques for T2I models.

Inference attacks have been widely studied and offer a systematic risk assessment of ML services; however, their implementation and the attack parameters for optimal estimation are challenging for non-experts. The emergence of advanced large language models presents a promising yet largely unexplored opportunity to develop autonomous agents as inference attack experts, helping address this challenge. In this paper, we propose AttackPilot, an autonomous agent capable of independently conducting inference attacks without human intervention. We evaluate it on 20 target services. The evaluation shows that our agent, using GPT-4o, achieves a 100.0% task completion rate and near-expert attack performance, with an average token cost of only $0.627 per run. The agent can also be powered by many other representative LLMs and can adaptively optimize its strategy under service constraints. We further perform trace analysis, demonstrating that design choices, such as a multi-agent framework and task-specific action spaces, effectively mitigate errors such as bad plans, inability to follow instructions, task context loss, and hallucinations. We anticipate that such agents could empower non-expert ML service providers, auditors, or regulators to systematically assess the risks of ML services without requiring deep domain expertise.

Large Language Models (LLMs) are highly effective for cybersecurity question answering (QA) but are difficult to deploy on edge devices due to their size. Quantization reduces memory and compute requirements but often degrades accuracy and increases vulnerability to adversarial attacks. We present EAGER, an edge-aligned defense framework that integrates parameter-efficient quantization with domain-specific preference alignment to jointly optimize efficiency, robustness, and accuracy. Unlike prior methods that address these aspects separately, EAGER leverages Quantized Low-Rank Adaptation (QLoRA) for low-cost fine-tuning and Direct Preference Optimization (DPO) on a self-constructed cybersecurity preference dataset, eliminating the need for human labels. Experiments show that EAGER reduces adversarial attack success rates by up to 7.3x and improves QA accuracy by up to 55% over state-of-the-art defenses, while achieving the lowest response latency on a Jetson Orin, demonstrating its practical edge deployment.

The science of science is an emerging field that studies the practice of science itself. We present the first study of the cybersecurity discipline from a science of science perspective. We examine the evolution of two comparable interdisciplinary communities in cybersecurity: the Symposium on Usable Privacy and Security (SOUPS) and Financial Cryptography and Data Security (FC).