Research data supporting "Harnessing Frontier LLMs for Automated Phishing-Resilient GRC Policy Generation in the Healthcare Sector: A Standards-Aligned, Human-in-the-Loop Approach"

This repository contains the complete research data supporting an Masters of Research in Cybersecurity dissertation evaluating three frontier large language models (GPT-5, Claude Opus 4.6 and Gemini 3.1 Pro) on their ability to generate phishing-resilient Governance, Risk and Compliance (GRC) policies for the UK NHS. It includes the nine generated policies (three models × three prompt types), the five-dimension scoring rubric with level descriptors and per-policy justifications, a 26-control standards-coverage matrix, and the verbatim prompt templates. Materials are provided to allow independent verification of the evaluation. Data collected May 2026.