Presented at EdMedia + Innovate Learning 2026, Edinburgh, and published in the proceedings.
Abstract
This paper documents a systematic framework for AI-human collaboration in educational worldbuilding, the practice of constructing coherent fictional universes that provide persistent narrative environments for pedagogical case studies. Developed through implementing seven interconnected cases in undergraduate cybersecurity policy education, our methodology comprises four integrated components: structured prompt engineering for scenario generation, YAML-based entity management for cross-case consistency, iterative human validation for pedagogical alignment, and reproducible technical infrastructure. Whereas traditional case studies typically present bounded, isolated scenarios, our approach treats AI as constrained collaborative partner operating within explicit pedagogical boundaries to generate persistent narrative environments. Implementation across three semesters with undergraduate and graduate students (n=110) suggested distinctive engagement patterns: spontaneous recognition of recurring entities, cross-scenario analytical synthesis, and more nuanced ethical reasoning than observed with traditional isolated cases. Iterative co-creation with AI, supported by systematic quality control, maintained fictional universe integrity and representational appropriateness throughout development. We discuss transferability beyond cybersecurity contexts and offer a reproducible methodology educators may adapt to their own educational settings.
Published in the Proceedings of EdMedia + Innovate Learning 2026 (AACE). https://www.learntechlib.org/p/2129766/
Reuse
Citation
@inproceedings{straight2026,
author = {Straight, Ryan and Straight, Ryan},
publisher = {Association for the Advancement of Computing in Education
(AACE)},
title = {Systematic {AI-Human} {Collaboration} for {Worldbuilding} in
{Cybersecurity} {Education}},
booktitle = {EdMedia + Innovate Learning 2026},
date = {2026},
url = {https://www.learntechlib.org/p/2129766/},
langid = {en}
}