Fine-Tuning AI to Assist in Building Curriculum for the CIA Triad and Cyber Kill Chain

Abstract

Educators face the continuous challenge of updating their teaching materials with the fast-paced changes in cybersecurity and integrating emerging topics into their existing content. This work presents the development progress of a cybersecurity curriculum fine-tuned Large Language Model (LLM) designed to assist educators in creating engaging curricular materials for introductory cybersecurity topics. Recent advancements in Natural Language Processing (NLP) and the increased number of openly available LLMs like OpenAI's GPT and Meta's Llama present an opportunity to fine-tune these models for specific domains like cybersecurity. The fine-tuned LLMs offer a potential solution by reducing the amount of work necessary to regularly update curricular content, including lab assignments, assessments, in-class activities, and other additional resources for both synchronous and asynchronous classrooms.

Ähnliche Arbeiten