Cybersecurity Capture-the-Flag (CTF) tournaments are well-understood to teach skills necessary for success in today's cybersecurity field. However, that does not mean CTFs are without critique. In the age of Generative Artificial Intelligence (AI), particularly for large-scale CTF tournaments, the use of Generative AI may be permitted or even encouraged to match the reality of today's practitioners, who are using AI systems to protect data, systems, and people. In such a situation, CTF participants must themselves balance the use of Generative AI with its overuse--effectively self-police the draw to offload one's thinking to the machine in pursuit of correct answers and prizes. In this paper, we introduce a framework for combating the overuse of Generative AI in cybersecurity educational activities through the building our own CTF using Augmented Reality (AR) technologies. Written by the nineteen undergraduate students who developed the CTF along with our professor who supervised our work, we argue that using the pedagogic lenses of the "see one, do one, teach one" model and peer learning allowed us to reinterpret our efforts on our CTF into a vision of shared labor and shared responsibility. This reframing of our own understanding of our work effectively acted as a counterbalance, keeping us focused on using Generative AI as a tool and not a crutch, leading to improved educational outcomes for us as individuals and the group as a whole. We hope that documenting our experiences inspires others to adopt similar counterbalance techniques where and when appropriate.

Keywords: augmented reality, capture-the-flag, cybersecurity, pedagogy, "see one do one teach one" model, peer learning

Rapid advances in AI, automation, and hyperconnectivity are outpacing human habits, producing pervasive ambiguity. Drawing on John Dewey's philosophy of habit as growth through disruption and inquiry, this paper reconceptualizes cybersecurity education as cultivating reflective, cyber-attuned habits across society—not only training specialists. Dewey's account of growth through disruption, inquiry, and reorganization are translated into three educational design moves: (1) embed reflective inquiry within procedural exercises; (2) employ inquiry-based, experiential formats (e.g., capture-the-flag, cyber-defense exercises, cyber-ranges) to practice reasoning under uncertainty; and (3) extend learning to social practices of verification and shared deliberation beyond technical settings. The approach turns error into material for growth and equips learners to act with intelligent adaptability.

Keywords: cybersecurity education, John Dewey’s philosophy, habit, growth

Cybersecurity awareness training is central to education and practice, yet persistent human error continues to expose organizations to breaches. AI-enabled attacks such as deepfakes, voice-cloned vishing, and automated spear phishing make these vulnerabilities even more consequential. This systematic review synthesizes 26 studies (2008–2025) using varied designs and training formats, from gamified learning and face-to-face sessions to e-learning, nudges, and simulated phishing. We introduced a residual-risk framework to capture outcomes that traditional effectiveness measures overlook. Residual Insecure Behavior (RIB) reflects the percentage of participants who continued risky practices after training, while Residual Knowledge Gap (RKG) indicates knowledge deficits that persisted. Across studies, improvements were common, but residual risks remained significant with phishing susceptibility often exceeding 10%, and knowledge gaps frequently surpassing 30%. Gamified approaches showed stronger behavioral effects, while conventional methods often raised awareness but left large gaps. For educators, these findings underscore that statistical gains can mask enduring weaknesses. By teaching and applying RIB and RKG, instructors can help students, practitioners, and organizations focus not just on learning outcomes, but on reducing real-world exposure in an AI-driven threat landscape.

Keywords: cybersecurity awareness, residual risk, insecure behavior, training effectiveness

The rapid growth of cloud and hybrid computing has brought significant scale, complexity, and security challenges to IT operations. Traditional rule-based monitoring systems and signature-based Security Information and Event Management (SIEM) tools are no longer sufficient to process the enormous volume of events generated in modern environments or to provide timely, accurate detection of incidents. Artificial Intelligence for IT Operations (AIOps) has emerged as a transformative approach by combining machine learning, predictive modeling, big data analytics, and automation to improve anomaly detection, optimize resource allocation, and accelerate the process of identifying root causes. Empirical studies report that AIOps platforms can reduce mean time to detection by nearly half and cut audit preparation time by up to 60%, underscoring their advantages over conventional methods. In addition to performance monitoring, AIOps is increasingly applied to security and compliance, enabling automated evidence collection, support for zero-trust architectures, and AI-assisted remediation workflows. Despite these benefits, reliance on opaque "black-box" models raises concerns around explainability, accountability, and regulatory compliance, particularly in mission-critical domains. Multi-cloud and hybrid infrastructures further complicate deployment due to interoperability issues, data silos, and risks of algorithmic bias. This paper reviews academic and industry work on AI-driven cloud security and operations from 2022 to 2025, outlines a taxonomy of AIOps functions spanning detection, compliance, response, and governance, and identifies unresolved challenges such as adversarial resilience, transparency, and multi-cloud coordination. Finally, future directions are discussed, including explainable and neuro-symbolic AIOps, federated analytics for distributed environments, and autonomous self-healing infrastructures. The review aims to provide researchers and practitioners with a consolidated reference for developing trustworthy, scalable, and secure AI-driven cloud operations.

Keywords: Artificial Intelligence for IT Operations (AIOps), Cloud Security, Hybrid Cloud, Multi-Cloud, Predictive Analytics, Anomaly Detection, Compliance Automation, Self-Healing Systems, Zero Trust, Workflow Orchestration, Explainable AI (XAI), Adversarial Machine Learning, Federated Analytics, IT Service Management, Big Data Analytics

Large Language Models (LLMs) have proven beneficial in aiding student learning across a multitude of domains such as computer science, data science, and mathematics. While these chatbots show promise, they can be impractical to deploy in situations where quality student data is not available.

Hint generation for cybersecurity can be feasible with the technology we have today if we make the right simplifications. First, we need human-in-the-loop systems because the training data used to train general LLMs may not cover cybersecurity well. Second, using modular agents allows us to access dynamic data that is outside the training dataset and add specificity to the hints.

In this paper, we leverage n8n, an agent deployment service, to establish the connections between our agents and Discord, the messaging system that our classroom uses to offer students a streamlined learning experience when working on interactive cybersecurity exercises. We have tested this in the classroom and have shown that it is feasible.

Keywords: Cybersecurity, Cybersecurity Education, Generative AI, Retrieval Augmented Generation

As teenagers increasingly engage with digital technology, cybersecurity vulnerabilities present significant risks to their online safety and privacy. Adolescents who lack awareness of secure online practices are particularly vulnerable to malicious actors seeking to exploit them. This paper investigates the relationship between real-world online behavior of teenagers, cybersecurity risks, and device interactions. The primary data set used for this analysis is Teenage online behavior and cybersecurity risks. First, we consider demographic information: age, education, time spent online. We correlate this with online behaviors: use of a VPN, type of equipment (computer, mobile), use of public internet, engagement with risky websites. Finally, we analyze the data set using a combination of demographic and behavioral patterns to search for high-risk, negative outcomes. For our research, we analyze teenage online behavior patterns to identify key risk factors, develop predictive models for cybersecurity vulnerabilities, and produce actionable visualizations that illustrate the relationship between digital literacy and online safety. Our findings utilize data and business analytics to provide evidence-based recommendations for parents, educators, and policymakers to enhance teenage cybersecurity awareness and protective strategies.

Keywords: online behavior, teenagers, adolescent, privacy, cybersecurity risk, data analysis

Security convergence, the integration of cybersecurity and physical security, has been discussed for over two decades, yet organizations still face challenges in defining and implementing effective strategies. This article explores the evolution of security convergence, highlighting key overlaps between physical and virtual security, especially with the rise of computerized operations in critical infrastructure. Through a 2023 survey and in-depth interviews with security professionals, four best practices are identified: implementing employee risk ratings, utilizing decision matrices, establishing fusion centers, and fostering a supportive organizational culture. These practices enhance collaboration and optimize security operations, demonstrating that effective convergence is not just about structural integration but also strategic coordination and cultural alignment for improved organizational resilience.

Keywords: Security convergence, Cybersecurity and physical security integration enterprise risk management, fusion centers, Security governance, workforce risk rating

The rapid digitalization of nuclear power plants (NPPs) and the deployment of advanced and small modular reactors (A/SMRs) have expanded the cybersecurity attack surface within the nuclear sector. This evolution introduces unique challenges beyond those faced in general information technology (IT), operational technology (OT) and industrial control system (ICS) security, due to nuclear power’s regulatory rigor, safety-critical nature, and operational needs.

A pressing workforce gap persists; cybersecurity graduates typically lack nuclear-specific context and retraining them for industry readiness requires 12–18 months, creating a significant burden. This paper addresses this gap by defining the domains of knowledge that nuclear cybersecurity specialists must master, spanning cybersecurity, nuclear engineering, OT/ICS security, and regulatory governance. We propose a curricular framework integrating technical, regulatory, and applied learning components to accelerate workforce readiness.

Our approach builds on existing findings that current curricula inadequately integrate nuclear engineering and cybersecurity, shifting the discourse from why specialization is needed to what knowledge must be taught. The recommendations have implications for workforce development and long-term resilience of the nuclear energy sector.

Keywords: nuclear cybersecurity, workforce development, higher education, OT/IT cybersecurity

Large Language Models (LLMs) are increasingly explored as tools for software development and could further constitute a supplementary source for the development of varied examples intended for pedagogical use. While they can improve productivity, their ability to produce code that is both secure and compliant with Secure Software Development (SSD) practices remains uncertain, raising concerns about their role in cybersecurity education. If LLMs are to be integrated effectively, students must be trained to critically evaluate generated code for correctness and vulnerabilities, raising an important question: How can LLM-generated code be effectively and securely incorporated into Cybersecurity education for teaching vulnerability analysis? This paper introduces CodeWars, a novel teaching methodology that combines LLM-generated and human-written code to examine how students engage with vulnerability detection tasks. CodeWars was implemented as a pilot study with a total of 32 students at Cardiff University and the University of Waikato, where students analyzed flawed, secure, and mixed-origin code samples. By comparing student approaches, analysis, and perceptions, the study provides insights into how vulnerabilities are detected, how code origins are distinguished, and how SSD practices are applied. Our analysis of student feedback and interviews indicates that Codewars produced structured and accessible code, simplifying vulnerability identification and offering educators the means to efficiently develop varied SSD teaching applications. These findings illuminate both the advantages and constraints of employing LLMs in secure coding and position this study as a foundational step toward the responsible adoption of AI in Cybersecurity education.

Keywords: Cybersecurity Education, Vulnerability analysis, Secure Software Development, Large Language Models, Cyber-security Pedagogy, GenAI

The rapid development of artificial intelligence, including agents and deepfake techniques, has accelerated phishing attacks and lowered the threshold for attackers. Modern phishing attacks now blend multiple tactics, including social engineering, URL spoofing, and AI deepfakes enabling adversaries to craft highly convincing messages that exploit human vulnerabilities and bypass traditional detection systems.

At the same time, current security awareness education struggles to keep up with the speed, sophistication, and complexity of these evolving threats.

To address this challenge, we propose a two-stage anti-phishing framework, CyberGLA, that combines technical defense and user-centered security education. In the Detection stage, we introduce EmailKnight, a spoof detection tool that performs multi-level email analysis.

Keywords: phishing detection, cybersecurity training, large language models, email spoofing, deepfake attacks, security awareness, human factors

Creating engaging cybersecurity education materials typically requires months of development time and specialized expertise. This paper describes how we used generative AI to address this challenge. We utilized Claude AI to generate a complete interactive platform that teaches students basic microelectronics through IoT hacking. Through iterative prompting, we generated more than 15,000 lines of functional code, including interactive visualizations, Python security tools, and gamified quizzes with real-time leaderboards. The curriculum guides students through the evolution of computing—from vacuum tubes to modern IoT devices—then helps them apply this foundation to discover real vulnerabilities. We implemented this platform at a GenCyber summer camp with 40 participants, where students identified actual security issues in AmpliPi audio systems—open-source network audio devices designed for multi-room audio distribution—including password weaknesses and denial of service flaws. The entire development process took only three weeks instead of the typical several months. The AI produced quality educational content, although we reviewed everything for technical accuracy and ethical considerations. During the camp, students remained engaged through competitive elements and hands-on labs, learning both theoretical concepts and practical skills. The students used AI-generated tools, including working implementations of SlowLoris and dictionary attacks, to test real systems. Our experience demonstrates that generative AI can efficiently create effective cybersecurity education materials that remain technically current. All materials are publicly available on GitHub for educational use. This approach could help educators stay on track with the rapidly evolving technology despite traditional curriculum development constraints.

Keywords: Generative AI, Cybersecurity Education, Interactive Learning, IoT Security, Hands-On Labs, Curriculum Development

This paper explores the evolving threat of deepfakes in the context of insider threats, particularly how advanced persistent threats (APTs) are leveraging AI-generated audio and video to impersonate job applicants and gain access to sensitive systems. While deepfakes have legitimate applications in entertainment, education, and business, they are increasingly being weaponized for deception and cyber intrusion. The paper outlines recent incidents, assesses technical vulnerabilities, and evaluates current risk management frameworks such as NIST RMF. It concludes with policy and technology recommendations to enhance detection and prevention strategies, especially during remote hiring and onboarding processes.

Keywords: deepfake, insider threat, artificial intelligence, AI, machine learning, ML, cybersecurity

AI is being interconnected with vital systems at an exponential rate, being described as the greatest shift in technology since the invention of the Internet. However, with the emergence of AI also involves the introduction of new critical vulnerabilities in the technology sector. This research will discuss the types of prompt injection attacks that AI can be subjected to, what they target and the possible repercussions of prompt injections. To counteract these attacks, solutions to detect different types of prompt injection will also be discussed, giving solutions to mitigate attacks that can expose critical data. Along with the solutions, different trade-offs between these solutions will be given. This research aims to expose the security issues that arise with the rapid implementation of experimental AI involving prompt injection and how to prevent it.

Keywords: artificial intelligence, AI, prompt injection, prompt mitigation, Ollama

Email on mobile has become a dominant communication channel for healthcare professionals, yet its constrained interface and context of use amplify vulnerability to social engineering attacks, especially phishing. This paper reports the development and empirical validation of the Healthcare Workers Phishing Risk Exposure (HWPRE) taxonomy, a 2×2 framework that positions individuals by (i) general email-phishing susceptibility; and (ii) ability to detect mobile-specific phishing cues. We followed a sequential three-phase design: (1) a Delphi study with cybersecurity subject matter experts to validate mobile-relevant phishing indicators and components of a susceptibility index; (2) a pilot to refine instruments and procedures; as well as (3) a large-scale study (N=300 healthcare workers) using scenario-based assessments on smartphone-generated email stimuli. We present the construction of the Healthcare Workers Email Phishing Susceptibility Index (HWEPSI), reliability/validity evidence, and statistical analyses relating HWPRE placement to role, experience, medical departments, prior training, and demographic indicators. The results show significant heterogeneity across departments and experience bands; in addition, the ability to recognize mobile cues does not follow uniformly with general susceptibility. We discuss implications for targeted Security Education, Training, and Awareness (SETA) programs and measurement-driven program evaluation. We conclude with practical guidance for integrating HWPRE into organizational phishing defense and directions for future research.

Keywords: Phishing, Social engineering, Healthcare cybersecurity, Mobile device cybersecurity, Human factors in cybersecurity, SETA in healthcare

This paper addresses a critical challenge facing cybersecurity educators: preparing students for AI-enhanced practice environments where effective action emerges from human-AI collaboration rather than individual expertise. Traditional instructional design frameworks assume human-centered learning processes that inadequately address distributed agency realities in contemporary cybersecurity operations. Drawing on Adams and Thompson’s posthuman inquiry methodology, this analysis develops a comprehensive pedagogical framework consisting of four principles: (1) Design for the Assemblage, Not the Individual, (2) Cultivate Relationality and Response-ability, (3) Embrace Emergence, Messiness, and Indeterminacy, and (4) Posthuman Assessment Approaches. The framework provides concrete instructional design implications, including strategies for configuring human-AI learning relations, integrating AI literacies across cognitive, civic, creative, and critical dimensions, and developing assemblage-aware cybersecurity case studies. These design implications bridge theoretical posthuman concepts with practical curriculum implementation through the lens of curriculum-as-lived rather than curriculum-as-plan. Preliminary implementation observations from an undergraduate cybersecurity ethics course demonstrate how posthuman-designed scenarios enable students to develop comfort with complexity and distributed analysis. Student reflections reveal progression from seeking singular solutions to embracing multiple valid perspectives, suggesting effective cultivation of human-AI collaborative competencies. The framework equips cybersecurity educators with both theoretical foundations and actionable design strategies for preparing students for distributed agency practice environments.

Keywords: distributed agency, artificial intelligence, cybersecurity education, posthumanism, human-AI collaboration, instructional design

EQiLevel is an emotionally adaptive AI tutoring system that integrates reinforcement learning (RL), large language models (LLMs), and real-time sentiment detection to personalize instruction dynamically. Traditional intelligent tutoring systems often follow rigid rules and fail to account for learners’ emotional states or individual learning preferences, reducing engagement. EQiLevel addresses this limitation by analyzing voice-based cues and adapting lesson difficulty, tone, and pacing in real time through a JSON-based Model Context Protocol (MCP). The MCP encodes emotion, performance, and learning style variables into structured state information, guiding GPT dialogue generation and reinforcement learning policy updates. Evaluation with simulated data demonstrated 78% successful adaptation to frustration cases, Whisper transcription accuracy with a 5.3% word error rate, emotion detection accuracy of 84% with 81% tone alignment, and improved RL convergence with average rewards rising from 0.41 to 0.63. In the context of cybersecurity education, EQiLevel illustrates how adaptive, emotion-aware tutoring can prepare learners to remain resilient under ambiguous and adversarial conditions, such as phishing awareness and threat analysis. By uniting technical adaptability with emotional intelligence, EQiLevel provides a scalable framework for inclusive, resilient, and effective cybersecurity education in the age of AI and automation.

Keywords: reinforcement learning, emotion-aware tutoring, cybersecurity education, adaptive learning, multimodal AI ambiguity resilience

Deepfakes are sophisticated, AI-generated alterations of images and videos that pose significant threats to cybersecurity, particularly with face-swapping techniques that can deceive and spread misinformation. A major limitation in current deepfake detection strategies is that the deepfakes used to train these models are often lower quality than those encountered in real-world scenarios [1]. This weakens model performance when tested against more sophisticated media alterations.

To bridge this gap, deepfake detection datasets must evolve to include high-quality deepfakes that better reflect the real-world threats. This study examines popular datasets such as Celeb-DF and DF-1.0, which revealed that despite efforts toward attribute variability, these datasets often lack demographic and facial variability. Even more important, many datasets do not publish attribute annotations, preventing researchers from fully understanding the imbalances present in their training data. Without dataset transparency, models may be unknowingly trained on skewed data, limiting their ability to generalize effectively.

The study presents an experiment using FaceSwap, a widely used deepfake creation tool, to investigate how dataset composition affects deepfake generation and detection. The experiment is dual-focused: (1) analyzing the impact of dataset augmentation through horizontal mirroring, which aims to increase facial orientation variability and improve model performance, and (2) evaluate how FaceSwap performs on subjects with different attributes to identify inherent skews in the deepfake generation process.

Keywords: Deepfake, Data Bias, AI Cybersecurity

As social networking sites (SNSs) have become integral to daily life, concerns about privacy and cybersecurity risks have intensified. Malicious actors exploit SNSs for phishing, malware distribution, and identity-driven attacks, often leveraging personal information voluntarily disclosed by users. This study investigates the relationships between SNS usage, personal information disclosure, cybersecurity behaviors, and experiences with cybersecurity threats. We employed a mixed-methods approach, combining survey data from 275 participants with semi-structured interviews. Correlation analyses revealed that frequency of SNS use and usage motivations---particularly for meeting new people and for self-presentation---were positively associated with higher levels of personal information disclosure. Disclosure of personal information and frequency of SNS usage were also significantly correlated with reported experiences of cybersecurity threats, though less so with protective cybersecurity behaviors. Interview responses highlighted both direct encounters with threats and broader perceptions of privacy vulnerabilities. Together, these findings underscore the complex interplay between social behavior on SNSs and cybersecurity risks, suggesting that greater user education and platform-level safeguards are necessary to mitigate emerging threats. We conclude with implications for cybersecurity awareness efforts and recommendations for future research.

Keywords: social networking sites, cybersecurity behavior, personal information disclosure, social engineering, privacy, mixed methods

This paper introduces an undergraduate course, Artificial Intelligence Applications in Cybersecurity, designed to equip students with Artificial Intelligence (AI) and Machine Learning (ML) skills to address modern cyber threats. The curriculum integrates supervised and unsupervised learning, deep learning, explainable AI (XAI), adversarial ML, and ethical considerations. Using accessible tools (Python, Google Colab) and real-world datasets (e.g., NSL-KDD, CICIDS2017, malware corpora), students complete phased projects progressing from classical ML baselines to deep learning with interpretability (SHAP/LIME) and robustness against adversarial attacks (FGSM/PGD with mitigation). The course aligns with data science and cybersecurity workforce frameworks, emphasizing reproducibility, communication, and responsible AI practices.

Keywords: Artificial Intelligence (AI), Cybersecurity Education, Machine Learning (ML), Explainable AI (XAI), Adversarial Machine Learning, Ethics and Fairness, Undergraduate Curriculum Design, Workforce Development

As cybersecurity becomes a cornerstone of global higher education, language has emerged as an unexpected point of vulnerability. Machine translation (MT) tools, increasingly used to render cybersecurity policies into multiple languages, often distort meaning by translating technical terms literally rather than conceptually. Words like firewall, phishing, or backdoor lose their intent in translation, creating barriers to comprehension and leaving multilingual learners at risk of misunderstanding critical policies. This paper explores the intersection of cybersecurity vocabulary, machine translation, and language equity, drawing on examples of mistranslations and language acquisition research to demonstrate how linguistic gaps can weaken both institutional safeguards and student confidence.

We argue that cybersecurity education must treat terminology with the same precision as code, recognizing that mistranslation not only undermines clarity but also compounds anxiety for multilingual learners navigating complex technical content. To address these challenges, the paper examines strategies such as the use of back-translation, custom glossaries, Universal Design for Learning (UDL) frameworks, and emerging AI translation tools like custom ChatGPT models. Together, these approaches highlight a pathway for higher education to balance inclusion with accuracy, ensuring that policies and coursework maintain both technical rigor and accessibility. By reframing cybersecurity not only as a technical field but also as a linguistic one, this research calls for a more intentional, equity-driven approach to translation that secures both data and learning outcomes.

Keywords: Cybersecurity education, Machine translation (MT), Multilingual learners, Universal Design for Learning (UDL), Artificial intelligence (AI) in translation, Custom ChatGPT, Higher education policy, Foreign Language Classroom Anxiety (FLCA)

The U.S. nuclear sector is undergoing rapid transformation, driven by the expansion of advanced reactors, digital modernization of legacy systems, and increasing interest in nuclear energy to meet AI-fueled energy demands. However, the cybersecurity talent pipeline is not keeping pace with this growth. This paper investigates the significant gap in nuclear cybersecurity education and proposes scalable strategies for colleges to address this critical need by promoting it as a viable and essential career path.

Through a multi-institutional landscape analysis of 16 cybersecurity and 12 nuclear engineering programs, we found that nuclear cybersecurity is largely absent from university curricula. Most students are unaware of the field’s existence, and few institutions offer hands-on training or interdisciplinary exposure. This lack of awareness leads to a shortage of specialized talent, forcing nuclear facilities to retrain generalist hires or rely on costly external consultants.

We present a framework for early pipeline cultivation grounded in Social Cognitive Career Theory and workforce development principles. Proposed solutions include student-led clubs, guest lectures, modular classroom kits, and summer boot camps. By increasing visibility and access to nuclear cyber content, we aim to break the self-reinforcing cycle of low awareness and limited specialization. This work underscores the critical role of education and advocacy in cultivating early interest and guiding students toward this emerging field. We call on academic institutions, national laboratories, and industry stakeholders to collaborate in establishing nuclear cybersecurity as a distinct and accessible career path within the broader cybersecurity and nuclear engineering ecosystems.

Keywords: nuclear cybersecurity, workforce development, curriculum, OT/IT cybersecurity

Cybersecurity threats continue to evolve in sophistication, increasingly targeting individuals as the weakest link in the security chain. While technical solutions remain essential, human-centered protective behaviors such as regular information backup are critical to mitigating risks from ransomware, device failure, and accidental deletion. This study investigates how past experiences with cyber incidents and awareness of threats influence backup practices among individuals in the United States. Drawing on Protection Motivation Theory (PMT), we conducted a mixed-methods survey (N{=}308) that measured threat appraisal (perceived severity, vulnerability) and coping appraisal (self-efficacy, response efficacy, response cost), along with threat awareness and prior experience. Multiple regression explained nearly half of the variance in backup behavior ($R^2{=}0.498$), with self-efficacy and threat awareness emerging as strong positive predictors; response cost was negative and significant. Qualitative responses illustrated experience- and awareness-driven adoption and highlighted common hybrid routines (cloud plus removable media) as well as barriers related to perceived effort or low perceived value. The findings underscore the importance of integrating human factors into cybersecurity programs and suggest concrete levers for awareness, design, and policy that reduce friction and promote routine backups.

Keywords: Cybersecurity, Protection Motivation Theory, Threat Awareness, Past Experience, Information Backup, Human Factors

Cyberattacks on critical infrastructures motivate a focus on cybersecurity awareness. A knowledge gap exists in the technical and non-technical understanding of cybersecurity, in the workforce. Closing this gap requires a multi-faceted approach -- of extreme importance is education. We use the NICE Workforce Framework TKS statements to develop a model of the most generalizable requirements needed to practice cybersecurity. We apply this model to increase cybersecurity language use and comprehension in all K-12 subjects, walking educators through a process incorporating cybersecurity into their lessons.

Keywords: Cybersecurity Education, K-12, Cybersecurity Literacy, NICE Workforce Framework

Cybersecurity awareness is widely recognised as an important requirement, but is frequently overlooked or addressed in ways that do not engage the interest of the target audience. In an attempt to broaden the options available for achieving this, the paper discusses the concept, design and evaluation of a new dice-based game designed to promote entry-level cyber security awareness in relation to common forms of attack and defence. The concept of the game is that players can defend against prior attacks, or use attacks to test defences, with the images on the different die faces denoting threats and safeguards of different strengths and which are countered in different ways. The discussion includes a worked example of the one of the game modes that has been designed in order to illustrate how players would take turns and make decisions in practice. It then presents initial results from a series of seven hands-on playtest sessions conducted with a range of audiences, including the general public, cyber educators and cyber professionals. The findings indicate that all audiences were positive about the game concept and found it enjoyable to play. Additionally, it was recognised to have value in raising and maintaining awareness, and would be a game that participants would play again and recommend to others.

Keywords: Cyber awareness, Cyber engagement, Dice, Gamification

Cybersecurity operations often involve repetitive tasks such as running Nmap scans, analyzing logs, and performing Open-Source Intelligence (OSINT) investigations. These processes are essential for maintaining security but consume time and resources that many small organizations cannot spare. While commercial automation platforms exist to reduce this workload, they are typically costly and inaccessible to businesses without dedicated IT staff. This paper investigates n8n, a self-hosted and low-cost workflow automation platform, as a practical alternative for cybersecurity automation. By integrating security tools and external large language models (LLMs) such as ChatGPT, Gemini, and Ollama, n8n can automate vulnerability scanning, assign severity ratings, and generate reports tailored to both technical and executive stakeholders. Experiments show that n8n workflows can effectively combine traditional scans with Artificial Intelligence (AI)-driven analysis to produce actionable outputs. Although limitations remain, including a steep learning curve and restrictions in the free tier, n8n demonstrates potential for broadening access to automation in cybersecurity. For small organizations, this approach provides a cost-effective way to strengthen security posture, while in academic contexts it provides a hands-on platform for teaching and experimenting with automation and AI in cybersecurity.

Keywords: Cybersecurity, Workflow Automation, n8n, Nmap, Artificial Intelligence (AI), Vulnerability scanning

Artificial Intelligence (AI) has become an increasingly powerful tool in various domains, particularly in image classification and object detection. As AI advances, novel methods to deceive machine learning models, such as adversarial patches, have emerged. These subtle modifications to images can lead to misclassification of objects, posing a substantial challenge to their reliability. In this paper, we present our research findings and literature on adversarial examples and object detection. This research builds upon the previous work by investigating the impact of small patches on object detection using YOLOv8. We started by exploring patterns within images and their influence on model accuracy. Then a follow-up study evaluating how adversarial patches, particularly those targeting animal patterns, affect YOLOv8’s ability to accurately detect objects. Additionally, we explore how untrained patterns impact the model’s performance, aiming to identify vulnerabilities and enhance the robustness of object detection systems.

Keywords: Artificial Intelligence, Object Detection, YOLOv8, Adversarial Patches, Machine Learning

This work introduces InfraLearn, a gamified learning platform designed to teach non-computer science students foundational background in cybersecurity for critical infrastructure. InfraLearn simulates attacks on a Distributed Energy Resource (DER) device, modeled after the Enphase Gateway solar monitor and implemented using a Flask-based API. Three prototype scenarios are developed: API spoofing, unauthorized remote shutdowns, and Living-off-the-Land (LoTL) downgrade exploitation. These scenarios are derived from real-world vulnerabilities in DER systems and integrated into a narrative-driven, web-based platform. Students interact with pre-configured virtual machines, guided code templates, and checkpoint quizzes, with optional AI support that reinforces comprehension while minimizing the need for prior programming experience. By situating cybersecurity concepts within the context of energy systems, InfraLearn makes abstract threats tangible and emphasizes the ethical application of defensive skills. This work demonstrates a scalable approach to engaging future engineers in securing critical infrastructure.

Keywords: Critical Infrastructure, Cybersecurity, Gamified Learning, Interactive

Cybersecurity education has traditionally emphasized reactive skills such as signature-based detection and log analysis. Although foundational, these approaches often overlook the need for contextual reasoning about user behavior and adversarial intent—skills critical in today’s evolving threat landscape. This paper presents a project-based educational framework that teaches students to build low-code, real-time data pipelines for intent-aware threat analytics, combining behavioral classification with explainable AI.

Simulated clickstream data from sandboxed or honeypot environments are analyzed to infer session intent, classify behaviors as benign or suspicious, and generate narrative summaries using large language models (LLMs). This layered approach helps learners understand not just what happened, but why—bridging raw telemetry and human reasoning through AI-generated explanations. Students engage in prompt engineering and evaluate explanation accuracy, fostering AI literacy alongside cybersecurity insight.

The framework emphasizes accessibility and modularity, using native cloud tools such as AWS Glue and Lambda that support deployment in both short workshops and semester-long courses. Instructors can customize classification heuristics, prompt structures, and visualization layers to suit curricular goals. Three core pedagogical objectives guide its design: fostering contextual reasoning, improving fluency in behavioral telemetry, and enhancing communication of AI-derived insights.

Early pilot demonstrations show increased student engagement, stronger critical thinking, and improved interpretation of behavioral indicators. By merging real-time analytics, generative AI, and project-based learning, this work redefines how cybersecurity learners approach intent detection—encouraging a shift from rote rule-following to interpretive, narrative-driven analysis of adversarial behavior.

Keywords: Behavioral Analytics, Clickstream Data, Cybersecurity Pedagogy, Explainable AI, Generative AI, Intent Detection, Large Language Models, Project-Based Learning, Real-Time Data Pipelines, Threat Interpretation

In recent years, the demand for practical, real-world cybersecurity education has grown dramatically. Traditional lecture- based methods often fall short in equipping students with the applied skills needed to detect, analyze, and respond to current cyber threats. This paper presents a project-based educational framework focused on the deployment, configuration, and use of real-world software such as Wazuh. Rather than following predetermined steps, students engage with realistic endpoint and network security scenarios, such as installing and configuring Wazuh agents, monitoring and interpreting live system and application logs, detecting simulated security incidents such as brute-force attacks and malware execution, and applying industry-aligned procedures. Evaluation of student performance demonstrates substantial improvements in alert interpretation, rule configuration, and application of cybersecurity knowledge. Our findings indicate that integrating Wazuh into coursework effectively develops both practical technical skills and analytical thinking, aligns with national workforce competency standards, and provides a model that other courses can adopt to integrate enterprise security tools into the classroom.

Keywords: Cybersecurity Education, Case Studies, Wazuh, Experiential Learning, Educational Framework

The rapid adoption of artificial intelligence across industries has outpaced security and privacy training for AI practitioners. This paper presents methods, modules, and findings from an experiential training program designed to address security and privacy challenges in AI systems development and deployment. We conducted two program iterations: a comprehensive 12-workshop series (May-October 2024) and a condensed 6-workshop format (January-February 2025). The program combined expert-led panel sessions with hands-on laboratory activities, engaging 78 participants from diverse professional backgrounds. Evaluation through pre-and post-evaluation surveys and qualitative observations revealed improvements in cybersecurity knowledge and AI security awareness.Participants demonstrated enhanced ability to identify vulnerabilities, implement security measures, and develop organizational policies for AI-related risk mitigation. The condensed format showed comparable learning outcomes with improved completion rates. This effort highlights the increased need to establish cybersecurity and privacy training for AI professionals to develop secure and trustworthy AI systems.

Keywords: AI security, cybersecurity training, privacy-preserving AI, experiential learning, professional development, adversarial machine learning

Cybersecurity risks are often treated as uniform, yet disparities across demographic groups suggest otherwise. This study investigates how ethnicity and geographic region shape cybersecurity outcomes in the United States, focusing on victimization, tool adoption, and awareness. Survey data from 470 adult participants were analyzed using ANOVA, Kruskal–Wallis tests, chi-square analyses, and logistic regression models. Results reveal two paradoxes. First, Asian/Pacific Islander respondents reported higher awareness and greater use of protective tools, yet also faced significantly elevated odds of identity theft, phishing losses, and account takeovers. Second, suburban residents exhibited higher preparedness than urban or rural populations, but consistently experienced greater exposure to cyber incidents, particularly financial fraud. Hispanic/Latinx and rural groups reported lower adoption of security tools, reflecting barriers of access and language. These findings highlight that awareness and adoption alone are insufficient when structural vulnerabilities and targeted exploitation are at play. The study underscores the need for culturally competent education, expanded infrastructure access, and adaptive monitoring systems to reduce disparities and promote a more equitable cybersecurity landscape.

Keywords: cybersecurity disparities, demographic factors, cyber victimization, digital divide, ethnicity, regional differences