Computing ethics and socially responsible computing (SRC) education has taken a prominent role in computing and technology programs in post-secondary education. Much of the prior work in this space describes the various approaches and strategies that scholars and educators have developed to teach ethics and SRC in technology and computing courses. There is limited work that explore and discuss how students within these courses describe the relationships and connections across various elements within technology ecosystems, such as computing technologies, users of these systems, and their impacts to various communities. Understanding students' perspectives when discussing the impacts of technology systems can give us a better sense of their own considerations when designing computing systems, as well as insights on how we can create learning designs that better support them in thinking about responsible computing. To explore this, I qualitatively analyzed students' responses to prompts that asked them to describe the pros and cons of different technology systems. I developed case studies that foreground themes within students' discussions and demonstrate how students articulated critical relationships across system and technology designs, their impacts, and various considerations within human-technology ecosystems. I also describe the implications of the work to learning designs that support students in thinking critically about the designs and impacts of technology, particularly in the use of narratives and speculative exploration in teaching socially responsible computing.
This study examines how B.S. ABET-accredited IT programs can align with industry needs while meeting the requirements of ABET standards. By analyzing current curricula against emerging technology trends (such as AI, cloud computing, and cybersecurity), we identify skill gaps and propose a flexible framework for enhancement. The study evaluates accredited programs and IT trends, revealing disparities in specialized domains. We recommend modular updates through electives and concentrations, offering institutions adaptable strategies to enhance graduate employability while maintaining accreditation.
This paper investigates the successes and challenges faced by under-graduate students in selecting topics for senior capstone projects in Computer Science. Conducted at California State University, Northridge over two semesters, the study compares instructor-driven (strict) and student-driven (dynamic) topic selection models. Survey responses and project evaluations from 64 students across 12 teams were analyzed to assess outcomes related to satisfaction, collaboration, and soft skills development. The findings indicate that student-driven topic selection promotes higher levels of motivation, ownership, and communication skills, though it introduces challenges around scoping, feasibility, and consistent evaluation. Instructor-driven models, while fostering academic rigor and deeper exposure to research-based work, may reduce student engagement and creative investment due to limited topic autonomy. These trade-offs highlight the value of a hybrid capstone structure, which blends student choice with structured faculty guidance to balance autonomy, mentorship, and educational outcomes.
Large language models (LLMs) are rapidly transforming programming education, yet little is known about how students' AI literacy and skills shape human-AI collaboration. We conducted an exploratory study with 23 computer science and robotics students who completed discipline-specific programming tasks in which the LLM produced the code and students guided and debugged it. AI literacy, measured with the SNAIL instrument, was generally high, with Practical Application rated highest and Critical Appraisal lowest, suggesting a risk of overconfidence. No robust links emerged between AI literacy, programming skill, task duration, or perceived chatbot output quality. Qualitative feedback revealed both enthusiasm for experimentation and frustration with AI errors. Findings indicate that brief, well-structured tasks can support productive AI-student collaboration, but stronger emphasis on critical evaluation, domain-specific skills, and appropriate scaffolding is needed.
Service-learning, an experiential learning pedagogy, has been shown to be effective in attracting and retaining students in computing, thereby contributing to the mission of broadening participation in computing (BPC). This paper presents a sustainable, project-based outreach model designed to engage students in educating communities about computing concepts. The model is grounded in a successful, multi-component service-learning program established at our undergraduate college to address the need to increase the number of students who persist in information technology (IT) and related majors.. Our unique program has proven to be sustainable through an established elective course and strong departmental support. This paper reviews the foundational components that have supported the outreach model over its 13 years of existence. More significantly, we highlight recent enhancements, including a new framework that unifies the format of our outreach activities and an online platform to document existing and new initiatives, facilitating adoption by others. The newly introduced outreach components include a series of career development events featuring IT alumni, as well as training workshops for K-12 teachers. We hope to inspire other institutions to adopt our model-or adapt its components-to enhance their own outreach initiatives and BPC efforts.
This paper presents Sonata, an AI-agent-assisted platform designed to address the challenges of educating students in secure quantumclassical hybrid networking. By integrating a multi-agent AI framework with a quantum-classical network simulator and a learning objects repository, Sonata offers a hands-on, interactive learning environment. The platform leverages large language models to provide conversational interfaces, personalized feedback, and scenario generation, enabling students to explore complex quantum networking concepts without requiring extensive prior expertise. The system includes a user-friendly GUI for designing network topologies, a Python API for programmatic control, and educational modules tailored to various knowledge levels. Sonata aims to bridge the gap in quantum education by making transdisciplinary concepts accessible, fostering experiential learning, and preparing a workforce proficient in quantum information sciences.
There is a continued need for computer and information technology (IT) professionals in industry. However, undergraduate IT programs often struggle to meet the demand for skilled workers. To improve career readiness of our IT majors, we organized a series of career workshop events featuring alumni with industry experience as panelists. In addition to a traditional panel discussion, the workshops incorporated a novel speed networking component, modeled after the speed dating concept, that enabled students to engage in direct, focused conversations with individual panelists. Pre-workshop survey responses revealed that students' expectations were broad and general, while their views post-workshop became more specific and targeted, showing that they became more informed about the career development process. Based on student feedback collected over four workshop iterations, we refined the event structure to maximize its impact. Overall, student feedback was very positive, indicating that these workshops effectively enhance career readiness and are a step forward in broadening participation in computing.
The global demand for cybersecurity professionals continues to outpace the available workforce, creating a critical gap that must be addressed through early education. This paper presents a projectbased teaching aid designed for secondary education to inspire students to learn cybersecurity in secondary school classrooms and pursue careers in cybersecurity. The teaching aid employs the Secure Software Development Life Cycle (SSDLC) as a lens to examine cyberattacks and includes three hands-on laboratory exercises that illustrate common cyberattacks and their countermeasures. While this paper focuses on three specific labs for illustration, the lab set is designed to be modular and expandable, enabling the integration of additional cybersecurity topics and attack scenarios. The project leverages free and open-source technologies, ensuring accessibility and equity by enabling web-based remote desktop connections to the lab infrastructure, making it easily implementable by secondary school educators. With most existing hands-on curricula designed for collegiate-level instruction and requiring significant technical setup, this work fills a crucial gap by providing age-appropriate and more accessible lab materials for secondary cybersecurity education.
As work begins on updating the ACM/IEEE-CS information technology curricula guidelines, this panel invites the IT education community to inform the development of the IT2027 curricular report. Shifts in the job market due to technological advancement and the fast-paced adoption of AI technology in all economic sectors call for reevaluating which competency domains we should retain, retire, or revise in the IT curricula. This panel will focus on revising the competency-based curricular framework with two innovations in mind: 1) include full educational pathways that extend from upper-level secondary education to encompass associate and baccalaureate IT degree programs, and 2) consider professional master's degrees in IT and computing bridge programs into graduate education. Another important focus of the panel is the integration and crosswalks between IT academic programs and the current and projected IT occupational landscape by using US and international classification systems for organizing and standardizing education and occupation data. Attendees will contribute to the early development of the IT2027 by sharing perspectives on future directions for IT education.
Modern software development is increasingly driven by cloudnative technologies, yet cybersecurity education has not kept pace. Traditional training environments often rely on static, pre-configured labs that fail to capture the complexity of real-world cloud infrastructures or adapt to individual learner needs. As a result, students frequently struggle to develop practical skills for securing dynamic environments such as containerized workloads, Continuous Integration/Continuous Deployment (CI/CD) pipelines, and Kubernetes clusters. To address these challenges, we present an AI-augmented cyber lab designed to deliver adaptive, hands-on security training. Our platform integrates Large Language Models (LLMs) for semantic analysis of student-submitted artifacts, including code, YAML configurations, and infrastructure-as-code policies, while a Reinforcement Learning (RL) agent dynamically adjusts the level of instructional support. The system is deployed on Kubernetes, leveraging container orchestration and automated threat simulation to replicate realistic attack scenarios. We validate the framework through a three-part evaluation methodology combining technical validation, system simulation, and expert review. The LLM Analyzer achieves an F1-score of 0.92 on a corpus of 1,500 real-world cloud-native artifacts, demonstrating high accuracy in identifying security misconfigurations. Simulation-based testing confirms the RL agent's adaptive behavior across diverse learner personas. Expert evaluation by five domain specialists yields average ratings of 4.4/5 for pedagogical value and confirms the system's potential for classroom deployment. This work establishes a rigorously validated, AI-enhanced architecture for cloud-native security education.
This study developed and validated an instrument to assess the quality of programming teachers in a face-to-face learning environment, known as the "Quality of Programming Teachers Inventory" (Q-PTI). The tool evaluates essential teaching competencies, including Delivery, Content, Personality, Pedagogy, Care, and Teaching Effectiveness. By systematically assessing these key qualities, the study provides a framework for evaluating IT educators and offers institutions a reliable instrument to assess teaching practices. The development process involved statistical methods, such as Factor Analysis and Cronbach's Alpha to ensure the validity and reliability of the instrument. Findings demonstrate that the Q-PTI tool can provide educators with actionable feedback, promoting professional development and enhancing student learning outcomes. This research highlights the crucial role of effective teaching in developing students' programming skills and preparing them for the rapidly evolving IT industry.
Differentiated Instruction is an inclusive teaching methodology that emphasizes skill development and the diversification of the learning process. Traditionally, this approach has been used in long-term educational settings and has not been easily applied to unconventional, immersive, and short-term experiences like coding camps. This paper unveils the European OSCAR initiative, which seeks to implement Differentiated Instruction in coding camps by linking them to school curricula and using a digital platform to streamline organization and assessment. This integration can enhance educational achievement and foster greater inclusivity.
The rise of generative artificial intelligence (GenAI) challenges the design, delivery, and integrity of assessments in undergraduate cybersecurity programs. While GenAI offers opportunities for enhancing instruction and learning, it also disrupts traditional written and authentic assessments, enabling students to bypass critical thinking and skill development. This paper examines institutional policies, pedagogical models, and assessment strategies to help instructors and programs adapt effectively. Drawing on recent studies, case examples, and curricular frameworks, the paper presents a rubric for addressing the use of GenAI by students in courses. These strategies can help maintain both the integrity and the relevance of cybersecurity education in a rapidly evolving technological context.
Experiential and collaborative activities-such as clubs, competitions, and advising-are proven to enhance student retention by fostering engagement and persistence. With shifting industry demands and workforce needs, competencies and professional dispositions have become essential. This experience report describes a collaboration between two community colleges in different regions to apply these methods in developing skills and a strong sense of belonging in computing. Activities were aligned with ACM/IEEE-CS/AAAI CS2023 guidelines, ensuring relevance to current industry standards. Shared practices yielded valuable insights into supporting student success and strengthening professional identity.
Recent years have witnessed a strong demand for cybersecurity professionals. The latest ABET accreditation guideline also lists cybersecurity as a required component in computer science curricula. Given the strong demand and accreditation requirement, it is important that the cybersecurity component, secure coding practices, can be well supported in computer science courses. Considering the importance of hands-on learning in cybersecurity education, and the established workflow of many computer science instructors in automatic grading of programming assignments, it will be the best of their interest if there is a tool that can seamlessly integrate cybersecurity hands-on learning into their existing auto grading workflow. In this work, we present CSEC, a platform for supporting CS students secure coding practices and improving their cybersecurity mindset. For instructors, CSEC provides auto code grading feature, and beyond that it offers auto assessment for secure coding/security exploit assignments and team-based secure programming projects that follow "build it, break it, and fix it" concept. This unique feature makes CSEC a versatile tool that integrates security exploit evaluation into the auto code grading workflow.
Artificial Intelligence (AI) has become embedded in critical sectors at a pace unmatched by previous technologies; however, its security has lagged significantly. This paper examines the widening gap between rapid AI adoption and the preparedness of systems, institutions, and practitioners to defend against novel attack surfaces-ranging from prompt injection and data poisoning to model theft, privacy leakage, and autonomous agent abuse. This paper presents empirical trends showing an accelerating rise in AIrelated incidents, analyzes systemic causes including educational gaps and a move-fast deployment culture, and documents highimpact failures from 2022-2025 that illustrate escalating real-world harm. On the basis of these findings, this paper argues that AI Security Education (AISecEd) should be elevated to a first-class priority in both AI/ML and cybersecurity curricula, with secure-bydesign principles embedded from the outset and dedicated courses to build specialized expertise. This paper maps AI course topics to security concepts, extends core cybersecurity modules to cover AI threats, and proposes a minimal baseline for industry adoption. By aligning education, practice, and policy around foundational AI security principles, the development community can close the gap between deployment and assurance-reducing catastrophic failures, strengthening trust, and ensuring that AI delivers on its promise without becoming a systemic liability.
As artificial intelligence (AI) increasingly transforms cybersecurity operations-from automated threat detection to real-time incident response-educational curricula must evolve to prepare graduates for this rapidly changing landscape. This panel will explore innovative strategies for integrating AI-driven security concepts, hands-on experiential learning, and ethical considerations into cybersecurity education. Bringing together academic thought leaders and curriculum experts, the discussion will highlight curricular adaptations, collaborative educational models, and assessment approaches designed to equip students with the skills and ethical foundation necessary to effectively address AI-fueled cyber threats.
As generative AI tools like ChatGPT become embedded in technical education, a critical challenge emerges: how can we ensure these tools foster learning rather than bypass it? This study provides empirical evidence that pedagogical design, not merely model access, determines the educational value of AI assistants. We developed a freely available custom conversational AI tool that embeds metacognitive scaffolding through structured prompts grounded in the Feynman Technique and learning science literature. In a quasi-experimental study within an undergraduate data structures course (N = 36), students using this structured AI assistant significantly outperformed peers using the same interface configured as a minimally prompted ChatGPT wrapper (92.7 vs. 74.3, p <.001). Gains were especially strong in abstraction, technical justification, and documentation, which are skills critical across software engineering, IT, and cybersecurity. These findings underscore a key insight: AI-integrated learning environments must be intentionally designed to prompt reflection, prediction, and explanation. By aligning AI interactions with evidence-based pedagogy, our framework demonstrates how to develop conceptual understanding, reduce automation bias, and support equitable learning outcomes as AI reshapes computing education.
This Lightning Talk presents preliminary research findings and introduces a new initiative aimed at bridging the gap between faculty expertise in cybersecurity and the career guidance provided by career counselors and academic advisors. As cybersecurity programs expand, many counselors and advisors, key influencers of student pathways, lack the knowledge, resources, and confidence to guide students toward cybersecurity education and careers. The Cybersecurity Career Awareness and Exploration Toolkit is a faculty-informed, DoD/NICE Cyber Workforce Frameworkaligned resource designed to help advisors and counselors discuss cybersecurity in ways that inspire students and highlight multiple entry points into the field. The toolkit builds directly on the results of a pilot Cybersecurity Career Awareness Workshop with 65 participants, which demonstrated 20-33 percentage point increases in knowledge of cybersecurity programs, certifications, and pathways, as well as more than 30-point gains in self-efficacy for advising students. Drawing from this evidence, the toolkit includes a student workbook, career pathway models and infographics, conversation guides, and student-friendly talking points. This talk will share the toolkit's vision, highlight pilot findings that validate its potential, and invite attendees to provide input on content development, evaluation strategies, and opportunities for collaborative dissemination. The ultimate goal is a collaborative, multi-institutional effort that strengthens the cybersecurity workforce pipeline by equipping advisors with effective, tested resources.
As cybersecurity threats grow more sophisticated-often using Artificial Intelligence (AI)-academic programs must evolve just as quickly to ensure workforce readiness. In response, the US National Science Foundation (NSF) funded a project to create curricular guidelines to focus on Cybersecurity and Artificial Intelligence. A novel part of the project was accelerating Knowledge Unit (KU) development using a custom AI assistant to aide in creating and curating the content of the KUs. This paper presents an analysis of the learning outcomes produced through this process, comparing their cognitive levels to five frameworks: the National Security Agency (NSA) Cyber Defense and Cyber Operations KU sets, ACM Computing Curricula 2023, ACM Cyber2Y, and ACM Cybersecurity guidelines (CSEC) 2017. In addition to the cognitive classification, a textual analysis was conducted to assess the similarity between AI-generated KU suggestions and the final versions of each KU. Our findings reveal that the AI-assisted knowledge units demonstrate cognitive rigor that is on par with-or in some cases exceeds-that of established frameworks, particularly in their alignment with Bloom's taxonomy verb usage for learning outcomes. Additionally, many learning outcomes in the final KU closely mirror the AI-generated suggestions. These results suggest that AI can act as an effective collaborator in curricular guideline design, contributing not only to efficiency but also to consistency and academic rigor.
