2026 ASEE Annual Conference & Exposition

Motivation
As higher education increasingly embraces flexible, technology-enhanced learning, understanding how different instructional formats affect engagement and learning is essential. While asynchronous courses provide accessibility and self-paced learning, they often struggle with reduced interaction and feedback compared to in-person settings. Integrating AI into asynchronous courses may offer a way to bridge engagement gaps through adaptive feedback and real-time scaffolding. This study compares two offerings of the same Cryptography and Data Security course—an asynchronous online section (14 students) taught in Fall 2025 and an in-person section (34 students) taught in Fall 2026—to examine how Generative AI/LLM tools influenced engagement, creativity, and development of content knowledge. The comparison seeks to provide evidence-based insights into how AI can enhance learning and inform future online and hybrid course designs in cybersecurity education.

Course Design and Activities
The asynchronous Cryptography and Data Security course is redesigned around three structured steps, which repeat in cycles of priming, discussions and reflections, and summarizing. Priming takes place first as students engage with course materials while accessing AI to support their comprehension and to make connections. Students then apply what they were primed to learn in discussions and reflections, students use AI to brainstorm creative projects, address challenges, and participate in peer-led discussion forums. While summarizing, students evaluate project strengths and weaknesses, prepare reports on their work, and reflect on their process using AI to scaffold their learning.
In contrast, the in-person offering of the same course follows a lecture-based format complemented by hands-on projects and assignments. This course design represents the balance in many active learning classes, emphasizing a combination of collaborative problem-solving and instructor-led demonstrations to reinforce theoretical concepts through practical application.

Assessment and Findings
Assessment of student learning in the two Cryptography and Data Security course offerings used multiple methods, including AI interaction logs, project rubrics, peer engagement measures, and self-reflections. Students in the asynchronous section used AI tools to clarify complex concepts, generate creative project ideas, and engage in deeper reflective discussions. Their projects demonstrated higher creativity, analytical depth, and stronger real-world relevance, while in-person students benefited from direct instructor interaction and collaboration, showing stronger procedural skills and teamwork. Survey data and engagement analytics revealed comparable conceptual understanding between groups, with the AI-supported asynchronous section reporting higher motivation, self-efficacy, and autonomy. Overall, the findings suggest that Generative AI/LLM integration can bridge engagement gaps in asynchronous formats while maintaining comparable learning outcomes.

Implications
Though piloted in cybersecurity, this three-step model is broadly transferable across disciplines, offering a scalable framework for reimagining asynchronous learning with AI-supported scaffolding, peer interaction, and creativity. The full paper expands on course design choices, project examples, comparative methodology, results, and future directions, offering evidence-based insights and a transferable framework for enhancing engagement and learning through AI integration across disciplines.