What is it about?
This paper presents CyberBOT, an intelligent chatbot designed to support reliable cybersecurity education. The system uses a method called retrieval-augmented generation (RAG), which allows it to draw information from course materials and generate contextually accurate answers. To ensure that responses are both correct and safe, CyberBOT employs a cybersecurity ontology—a structured knowledge framework that verifies the accuracy of generated answers. The tool has been deployed in a graduate-level cloud computing course at Arizona State University, helping students ask questions and receive validated, trustworthy explanations. This work demonstrates how artificial intelligence can be responsibly integrated into education, improving the reliability and alignment of AI-driven learning systems with academic curricula.
Featured Image
Photo by Igor Omilaev on Unsplash
Why is it important?
As large language models become increasingly integrated into education, ensuring their accuracy and trustworthiness has become a pressing concern—especially in technical and safety-critical fields such as cybersecurity. CyberBOT addresses this challenge by combining retrieval-augmented generation with an ontology-based validation framework, creating a system that not only generates informative answers but also verifies their correctness against authoritative domain knowledge. This dual-layer approach is both novel and timely, offering a safeguard against misinformation in AI-assisted learning. By demonstrating the successful deployment of an ontology-grounded chatbot in a real classroom setting, this work highlights a significant step toward developing dependable, curriculum-aligned AI tools that enhance both the quality and safety of modern digital education.
Perspectives
Working on this paper was an inspiring experience, as it brought together researchers from computing and education to address a common challenge—how to make artificial intelligence both reliable and educationally meaningful. Collaborating across disciplines allowed me to appreciate how technical innovation can directly support teaching and learning in real classrooms. I hope this work encourages others to explore how AI systems can be designed not only for efficiency or performance, but also for safety, trust, and pedagogical value. Most of all, I hope readers see this as an example of how responsible AI design can enhance human learning rather than replace it.
Chengshuai Zhao
Arizona State University
Read the Original
This page is a summary of: CyberBOT: Ontology-Grounded Retrieval Augmented Generation for Reliable Cybersecurity Education, November 2025, ACM (Association for Computing Machinery),
DOI: 10.1145/3746252.3761478.
You can read the full text:
Contributors
The following have contributed to this page
