AI Risk Assessor
Run a structured AI risk assessment across 60 scenarios spanning 10 risk families — Model, Data, Security, Governance, Human & Ethical, Operational, Monitoring, Business, Agentic AI, and Fail-Safe risks. Score likelihood and impact, apply control weights, and generate a prioritised risk register with heat map and export.
Config Editor
Tune the assessment framework to your organisation's risk appetite across five configuration tabs: Risk Matrix (5×5 rating bands), Answer Weights (evidence strength scoring), Survey Evidence (caps and coverage), Control Treatment (reduction factors), and Risk Scores (base score and strategy bias). Export and re-import configs as JSON.
User Guide
Complete reference documentation for the toolkit. Covers the full risk taxonomy with all 60 scenarios across 10 families, the scoring methodology (likelihood × impact, control weighting, evidence caps), how to interpret heat map output, maturity levels, and step-by-step instructions for running and exporting an assessment.
The Challenge
The story behind the toolkit — why AI-generated risk scores fail governance requirements, and why a deterministic, human-evidenced model is the right design choice. Covers six topics: the governance gap, the AI scoring dilemma, how the tool was built with AI, the prompting methodology, build cost, and what the finished product demonstrates.