arXiv subjects

Ashraf Matrawy

Publications and source records attributed to Ashraf Matrawy.

At least 19 recordsLinked to original sources

arXiv2026

Unanticipated Effects of Generative AI on Expertise Pathways and Performance Perception in System Administration

While industry discourse often emphasizes immediate productivity gains and frames GenAI primarily as a tool for automation, the integration of GenAI into system administration may involve deeper shifts in professional practice that are not yet fully understood. Drawing on 14 semi-structured interviews with IT professionals, this paper explores the lived reality of embedding GenAI into daily routines of troubleshooting, scripting, and system verification. Through inductive thematic analysis, we uncover two unanticipated socio-technical findings. First, we describe a "compression of traditional expertise pathways" where GenAI appears to function as both a mentor-like tutor and a "ladder-shortening" tool. While the tool can support faster task performance in unfamiliar domains, our findings suggest it may also reduce a practitioner's exposure to the foundational, hands-on cycles of building, failing, and debugging that historically served as the training ground for technical expertise. Second, we describe a "performance perception shift," where the speed of AI-assisted work begins to reset organizational and self-expectations for productivity. This shift may create a "two-speed culture" within teams and introduce "productivity guilt," as necessary manual work, even when required for safety or validation, is increasingly perceived as slow or a failure of efficiency. Our results raise broader questions about how GenAI may influence expertise development, how professional value is assessed in high-stakes technical environments, and the role of human judgment in complex technical environments.

cs.HC↗

arXiv2026

A No-Defense Defense Against Gradient-Based Adversarial Attacks on ML-NIDS: Is Less More?

Gradient-based adversarial attacks subtly manipulate inputs of Machine Learning (ML) models to induce incorrect predictions. This paper investigates whether careful architectural choices alone can yield an inherently robust Deep Neural Network (DNN)-based Network Intrusion Detection Systems (NIDS), without any additional explicit defenses. Through thousands of experiments, around 2200, varying network depth, feature dimensionality, activation functions, and dropout across FGSM, PGD, and BIM attacks, we show that shallower networks, reduced feature sets, and ReLU activation consistently and jointly reduce adversarial vulnerability. Moreover, a simple model following this recipe outperforms deeper, fully-featured adversarially trained models, while maintaining near-perfect clean-traffic detection and lower training times. Nevertheless, while less is more, the selection of the right less is what truly matters.

cs.LG↗

arXiv2026

Threat Modelling using Domain-Adapted Language Models: Empirical Evaluation and Insights

Large Language Models(LLMs) are increasingly explored for cybersecurity applications such as vulnerability detection. In the domain of threat modelling, prior work has primarily evaluated a number of general-purpose Large Language Models under limited prompting settings. In this study, we extend the research area of structured threat modelling by systematically evaluating domain-adapted language models of different sizes to their general counterparts. We use both LLMs and Small Language Models(SLMs) that were domain adapted to telecommunications and cybersecuirty. For the structured threat modelling, we selected the widely used STRIDE approach and the application area is 5G security. We present a comprehensive empirical evaluation using 52 different configurations (on 8 different language models) to analyze the impact of 1) domain adaptation, 2) model scale, 3) decoding strategies (greedy vs. stochastic sampling), and 4) prompting technique on STRIDE threat classification. Our results show that domain-adapted models do not consistently outperform their general-purpose counterparts, and decoding strategies significantly affect model behavior and output validity. They also show that while larger models generally achieve higher performance, these gains are neither consistent nor sufficient for reliable threat modelling. These findings highlight fundamental limitations of current LLMs for structured threat modelling tasks and suggest that improvements require more than additional training data or model scaling, motivating the need for incorporating more task-specific reasoning and stronger grounding in security concepts. We present insights on invalid outputs encountered and present suggestions for prompting tailored specifically for STRIDE threat modelling.

cs.CR↗

arXiv2025

SA-ADP: Sensitivity-Aware Adaptive Differential Privacy for Large Language Models

Despite advances in the use of large language models (LLMs) in downstream tasks, their ability to memorize information has raised privacy concerns. Therefore, protecting personally identifiable information (PII) during LLM training remains a fundamental challenge. Conventional methods like Differential Privacy-Stochastic Gradient Descent (DP-SGD) provide robust privacy protection via uniform noising, protecting PII regardless of its distinct sensitivity. This comes at the expense of the model's utility, leading to a trade-off. In this paper, we propose SA-ADP, a sensitivity-aware approach that allocates noise based on the sensitivity of individual PII. We evaluated our method on four datasets (ABCD, CUSTOMERSIM, Wikitext-2, and UNSW-NB15 ). Our results show that SA-ADP achieves results comparable to the baseline (No-DP) and the conventional DP-SGD. This means that our method did not degrade the model's utility while still maintaining strong privacy protection.

cs.LG↗

arXiv2025

Demystifying QUIC from the Specifications

QUIC is an advanced transport layer protocol whose ubiquity on the Internet is now very apparent. Importantly, QUIC fuels the next generation of web browsing: HTTP/3. QUIC is a stateful and connection oriented protocol which offers similar features (and more) to the combination of TCP and TLS. There are several difficulties which readers may encounter when learning about QUIC: i.) its rapid evolution (particularly, differentiation between the QUIC standard and the now deprecated Google QUIC), ii.) numerous RFCs whose organization, language, and detail may be challenging to the casual reader, and iii.) the nature of QUIC's cross-layer and privacy-centric implementation, making it impossible to understand or debug by looking at packets alone. For these reasons, the aim of this paper is to present QUIC in a complete yet approachable fashion, thereby demystifying the protocol from its specifications.

cs.NI↗

arXiv2025

PTMF: A Privacy Threat Modeling Framework for IoT with Expert-Driven Threat Propagation Analysis

Previous studies on PTA have focused on analyzing privacy threats based on the potential areas of occurrence and their likelihood of occurrence. However, an in-depth understanding of the threat actors involved, their actions, and the intentions that result in privacy threats is essential. In this paper, we present a novel Privacy Threat Model Framework (PTMF) that analyzes privacy threats through different phases. The PTMF development is motivated through the selected tactics from the MITRE ATT\&CK framework and techniques from the LINDDUN privacy threat model, making PTMF a privacy-centered framework. The proposed PTMF can be employed in various ways, including analyzing the activities of threat actors during privacy threats and assessing privacy risks in IoT systems, among others. In this paper, we conducted a user study on 12 privacy threats associated with IoT by developing a questionnaire based on PTMF and recruited experts from both industry and academia in the fields of security and privacy to gather their opinions. The collected data were analyzed and mapped to identify the threat actors involved in the identification of IoT users (IU) and the remaining 11 privacy threats. Our observation revealed the top three threat actors and the critical paths they used during the IU privacy threat, as well as the remaining 11 privacy threats. This study could provide a solid foundation for understanding how and where privacy measures can be proactively and effectively deployed in IoT systems to mitigate privacy threats based on the activities and intentions of threat actors within these systems.

cs.CR↗

arXiv2025

Exploring the Effect of DNN Depth on Adversarial Attacks in Network Intrusion Detection Systems

Adversarial attacks pose significant challenges to Machine Learning (ML) systems and especially Deep Neural Networks (DNNs) by subtly manipulating inputs to induce incorrect predictions. This paper investigates whether increasing the layer depth of deep neural networks affects their robustness against adversarial attacks in the Network Intrusion Detection System (NIDS) domain. We compare the adversarial robustness of various deep neural networks across both \ac{NIDS} and computer vision domains (the latter being widely used in adversarial attack experiments). Our experimental results reveal that in the NIDS domain, adding more layers does not necessarily improve their performance, yet it may actually significantly degrade their robustness against adversarial attacks. Conversely, in the computer vision domain, adding more layers exhibits a more modest impact on robustness. These findings can guide the development of robust neural networks for (NIDS) applications and highlight the unique characteristics of network security domains within the (ML) landscape.

cs.CR↗

arXiv2025

Evaluation Metrics for Misinformation Warning Interventions: Challenges and Prospects

Misinformation has become a widespread issue in the 21st century, impacting numerous areas of society and underscoring the need for effective intervention strategies. Among these strategies, user-centered interventions, such as warning systems, have shown promise in reducing the spread of misinformation. Many studies have used various metrics to evaluate the effectiveness of these warning interventions. However, no systematic review has thoroughly examined these metrics in all studies. This paper provides a comprehensive review of existing metrics for assessing the effectiveness of misinformation warnings, categorizing them into four main groups: behavioral impact, trust and credulity, usability, and cognitive and psychological effects. Through this review, we identify critical challenges in measuring the effectiveness of misinformation warnings, including inconsistent use of cognitive and attitudinal metrics, the lack of standardized metrics for affective and emotional impact, variations in user trust, and the need for more inclusive warning designs. We present an overview of these metrics and propose areas for future research.

cs.HC↗

arXiv2025

LLMs' Suitability for Network Security: A Case Study of STRIDE Threat Modeling

Artificial Intelligence (AI) is expected to be an integral part of next-generation AI-native 6G networks. With the prevalence of AI, researchers have identified numerous use cases of AI in network security. However, there are very few studies that analyze the suitability of Large Language Models (LLMs) in network security. To fill this gap, we examine the suitability of LLMs in network security, particularly with the case study of STRIDE threat modeling. We utilize four prompting techniques with five LLMs to perform STRIDE classification of 5G threats. From our evaluation results, we point out key findings and detailed insights along with the explanation of the possible underlying factors influencing the behavior of LLMs in the modeling of certain threats. The numerical results and the insights support the necessity for adjusting and fine-tuning LLMs for network security use cases.

cs.CR↗

arXiv2025

A Datagram Extension to DNS over QUIC: Proven Resource Conservation in the Internet of Things

In this paper, we investigate the Domain Name System (DNS) over QUIC (DoQ) and propose a non-disruptive extension, which can greatly reduce DoQ's resource consumption. This extension can benefit all DNS clients - especially Internet of Things (IoT) devices. This is important because even resource-constrained IoT devices can generate dozens of DNS requests every hour. DNS is a crucial service that correlates IP addresses and domain names. It is traditionally sent as plain-text, favoring low-latency results over security and privacy. The repercussion of this can be eavesdropping and information leakage about IoT devices. To address these concerns, the newest and most promising solution is DoQ. QUIC offers features similar to TCP and TLS while also supporting early data delivery and stream multiplexing. DoQ's specification requires that DNS exchanges occur over independent streams in a long-lived QUIC connection. Our hypothesis is that due to DNS's typically high transaction volume, managing QUIC streams may be overly resource intensive for IoT devices. Therefore, we have designed and implemented a data delivery mode for DoQ using QUIC datagrams, which we believe to be more preferable than stream-based delivery. To test our theory, we analyzed the memory, CPU, signaling, power, and time of each DoQ delivery mode in a setup generating real queries and network traffic. Our novel datagram-based delivery mode proved to be decisively more resource-friendly with little compromise in terms of functionality or performance. Furthermore, our paper is the first to investigate multiple queries over DoQ, to our knowledge.

cs.NI↗

arXiv2025

strideSEA: A STRIDE-centric Security Evaluation Approach

Microsoft's STRIDE methodology is at the forefront of threat modeling, supporting the increasingly critical quality attribute of security in software-intensive systems. However, in a comprehensive security evaluation process, the general consensus is that the STRIDE classification is only useful for threat elicitation, isolating threat modeling from the other security evaluation activities involved in a secure software development life cycle (SDLC). We present strideSEA, a STRIDE-centric Security Evaluation Approach that integrates STRIDE as the central classification scheme into the security activities of threat modeling, attack scenario analysis, risk analysis, and countermeasure recommendation that are conducted alongside software engineering activities in secure SDLCs. The application of strideSEA is demonstrated in a real-world online immunization system case study. Using STRIDE as a single unifying thread, we bind existing security evaluation approaches in the four security activities of strideSEA to analyze (1) threats using Microsoft threat modeling tool, (2) attack scenarios using attack trees, (3) systemic risk using NASA's defect detection and prevention (DDP) technique, and (4) recommend countermeasures based on their effectiveness in reducing the most critical risks using DDP. The results include a detailed quantitative assessment of the security of the online immunization system with a clear definition of the role and advantages of integrating STRIDE in the evaluation process. Overall, the unified approach in strideSEA enables a more structured security evaluation process, allowing easier identification and recommendation of countermeasures, thus supporting the security requirements and eliciting design considerations, informing the software development life cycle of future software-based information systems.

cs.CR↗

arXiv2025

Privacy Preservation Techniques (PPTs) in IoT Systems: A Scoping Review and Future Directions

Privacy preservation in Internet of Things (IoT) systems requires the use of privacy-enhancing technologies (PETs) built from innovative technologies such as cryptography and artificial intelligence (AI) to create techniques called privacy preservation techniques (PPTs). These PPTs achieve various privacy goals and address different privacy concerns by mitigating potential privacy threats within IoT systems. This study carried out a scoping review of different types of PPTs used in previous research works on IoT systems between 2010 and early 2023 to further explore the advantages of privacy preservation in these systems. This scoping review looks at privacy goals, possible technologies used for building PET, the integration of PPTs into the computing layer of the IoT architecture, different IoT applications in which PPTs are deployed, and the different privacy types addressed by these techniques within IoT systems. Key findings, such as the prominent privacy goal and privacy type in IoT, are discussed in this survey, along with identified research gaps that could inform future endeavors in privacy research and benefit the privacy research community and other stakeholders in IoT systems.

cs.CR↗

arXiv2024

A Novel Perturb-ability Score to Mitigate Evasion Adversarial Attacks on Flow-Based ML-NIDS

As network security threats evolve, safeguarding flow-based Machine Learning (ML)-based Network Intrusion Detection Systems (NIDS) from evasion adversarial attacks is crucial. This paper introduces the notion of feature perturb-ability and presents a novel Perturb-ability Score (PS), which quantifies how susceptible NIDS features are to manipulation in the problem-space by an attacker. PS thereby identifies features structurally resistant to evasion attacks in flow-based ML-NIDS due to the semantics of network traffic fields, as these features are constrained by domain-specific limitations and correlations. Consequently, attempts to manipulate such features would likely either compromise the attack's malicious functionality, render the traffic invalid for processing, or potentially both outcomes simultaneously. We introduce and demonstrate the effectiveness of our PS-enabled defenses, PS-guided feature selection and PS-guided feature masking, in enhancing flow-based NIDS resilience. Experimental results across various ML-based NIDS models and public datasets show that discarding or masking highly manipulatable features (high-PS features) can maintain solid detection performance while significantly reducing vulnerability to evasion adversarial attacks. Our findings confirm that PS effectively identifies flow-based NIDS features susceptible to problem-space perturbations. This novel approach leverages problem-space NIDS domain constraints as lightweight universal defense mechanisms against evasion adversarial attacks targeting flow-based ML-NIDS.

cs.CR↗

arXiv2024

Cybersecurity Challenge Analysis of Work-from-Anywhere (WFA) and Recommendations guided by a User Study

Many organizations were forced to quickly transition to the work-from-anywhere (WFA) model as a necessity to continue with their operations and remain in business despite the restrictions imposed during the COVID-19 pandemic. Many decisions were made in a rush, and cybersecurity decency tools were not in place to support this transition. In this paper, we first attempt to uncover some challenges and implications related to the cybersecurity of the WFA model. Secondly, we conducted an online user study to investigate the readiness and cybersecurity awareness of employers and their employees who shifted to work remotely from anywhere. The user study questionnaire addressed different resilience perspectives of individuals and organizations. The collected data includes 45 responses from remotely working employees of different organizational types: universities, government, private, and non-profit organizations. Despite the importance of security training and guidelines, it was surprising that many participants had not received them. A robust communication strategy is necessary to ensure that employees are informed and updated on security incidents that the organization encounters. Additionally, there is an increased need to pay attention to the security-related attributes of employees, such as their behavior, awareness, and compliance. Finally, we outlined best practice recommendations and mitigation tips guided by the study results to help individuals and organizations resist cybercrime and fraud and mitigate WFA-related cybersecurity risks.

cs.CR↗

arXiv2024

Introducing Adaptive Continuous Adversarial Training (ACAT) to Enhance ML Robustness

Adversarial training enhances the robustness of Machine Learning (ML) models against adversarial attacks. However, obtaining labeled training and adversarial training data in network/cybersecurity domains is challenging and costly. Therefore, this letter introduces Adaptive Continuous Adversarial Training (ACAT), a method that integrates adversarial training samples into the model during continuous learning sessions using real-world detected adversarial data. Experimental results with a SPAM detection dataset demonstrate that ACAT reduces the time required for adversarial sample detection compared to traditional processes. Moreover, the accuracy of the under-attack ML-based SPAM filter increased from 69% to over 88% after just three retraining sessions.

cs.LG↗

arXiv2024

Privacy Engineering in Smart Home (SH) Systems: A Comprehensive Privacy Threat Analysis and Risk Management Approach

Addressing trust concerns in Smart Home (SH) systems is imperative due to the limited study on preservation approaches that focus on analyzing and evaluating privacy threats for effective risk management. While most research focuses primarily on user privacy, device data privacy, especially identity privacy, is almost neglected, which can significantly impact overall user privacy within the SH system. To this end, our study incorporates privacy engineering (PE) principles in the SH system that consider user and device data privacy. We start with a comprehensive reference model for a typical SH system. Based on the initial stage of LINDDUN PRO for the PE framework, we present a data flow diagram (DFD) based on a typical SH reference model to better understand SH system operations. To identify potential areas of privacy threat and perform a privacy threat analysis (PTA), we employ the LINDDUN PRO threat model. Then, a privacy impact assessment (PIA) was carried out to implement privacy risk management by prioritizing privacy threats based on their likelihood of occurrence and potential consequences. Finally, we suggest possible privacy enhancement techniques (PETs) that can mitigate some of these threats. The study aims to elucidate the main threats to privacy, associated risks, and effective prioritization of privacy control in SH systems. The outcomes of this study are expected to benefit SH stakeholders, including vendors, cloud providers, users, researchers, and regulatory bodies in the SH systems domain.

cs.CR↗

arXiv2023

Evasion Adversarial Attacks Remain Impractical Against ML-based Network Intrusion Detection Systems, Especially Dynamic Ones

Machine Learning (ML) has become pervasive, and its deployment in Network Intrusion Detection Systems (NIDS) is inevitable due to its automated nature and high accuracy compared to traditional models in processing and classifying large volumes of data. However, ML has been found to have several flaws, most importantly, adversarial attacks, which aim to trick ML models into producing faulty predictions. While most adversarial attack research focuses on computer vision datasets, recent studies have explored the suitability of these attacks against ML-based network security entities, especially NIDS, due to the wide difference between different domains regarding the generation of adversarial attacks. To further explore the practicality of adversarial attacks against ML-based NIDS in-depth, this paper presents several key contributions: identifying numerous practicality issues for evasion adversarial attacks on ML-NIDS using an attack tree threat model, introducing a taxonomy of practicality issues associated with adversarial attacks against ML-based NIDS, identifying specific leaf nodes in our attack tree that demonstrate some practicality for real-world implementation and conducting a comprehensive review and exploration of these potentially viable attack approaches, and investigating how the dynamicity of real-world ML models affects evasion adversarial attacks against NIDS. Our experiments indicate that continuous re-training, even without adversarial training, can reduce the effectiveness of adversarial attacks. While adversarial attacks can compromise ML-based NIDSs, our aim is to highlight the significant gap between research and real-world practicality in this domain, which warrants attention.

cs.CR↗

arXiv2023

A Security-aware Network Function Sharing Model for 5G Slicing

Sharing Virtualized Network Functions (VNFs) among different slices in Fifth Generation (5G) is a potential strategy to simplify the system implementation and utilize 5G resources efficiently. In this paper, we propose a security-aware VNF sharing model for 5G networks. The proposed optimization model satisfies the service requirements of various slices, enhances slice security by isolating their critical VNFs, and enhances resource utilization of the underlying physical infrastructure. The model tries to systematically decide on sharing a particular VNF based on two groups of constraints; the first group of constraints is common assignment constraints used in the existing literature. The second group is the novel security constraints that we propose in this work; the maximum traffic allowed to be processed by the VNF and the exposure of the VNF to procedures sourced via untrusted users or access networks. This sharing problem is formalized to allow for procedure-level modeling that satisfies the requirements of slice requests in 5G systems. The model is tested using standard VNFs and procedures of the 5G system rather than generic ones. The numerical results of the model show the benefits and costs of applying the security constraints along with the network performance in terms of different metrics.

cs.NI↗