A structured self-assessment for organisations deploying AI in public-facing or institutional settings.
This checklist measures governance readiness — not AI quality.
A high score does not mean an AI system is accurate, unbiased, lawful, safe, or appropriate for deployment. A low score does not, by itself, mean a system should be discontinued. It asks a narrower question: does the organisation have demonstrable governance mechanisms around the system’s authority, accountability, discretion, transparency, public value, ongoing oversight, and — where relevant — its vendors?
This checklist is built around HCI’s four-part governance framework — Legitimate Authority, Real Accountability, Bounded Discretion, and Public Value — four principles adapted from public-administration thinking and applied to AI systems instead of human institutions. These are the same governance dimensions that structure HCI’s full Ethical Expertise review, simplified here into a binary self-assessment. Three further sections add the operational scaffolding those principles depend on in practice.
How to use this: work through it in order. For each item, mark Yes, No, or N/A. Check Yes only when you can point to evidence — a policy that exists only as a draft, a process nobody actually follows, or a control that depends on one person’s informal knowledge should not be counted as operational. N/A items are excluded from your score; they don’t count in your favour.
0Define the System
Before assessing governance, establish exactly what’s being assessed. This section isn’t scored — it’s context for everything that follows. Optional, but useful if you email yourself a copy of your results.
0 of 35 answered
Please answer every item before calculating — scroll up to the highlighted question.
Your Governance Readiness
This score does not measure whether the AI system itself is ethical, lawful, safe, or fit for deployment. It measures whether the organisation has demonstrable governance mechanisms in place around it. This checklist relies on your own answers — it does not independently verify the documents, processes, decisions, or evidence behind them. HCI’s Ethical Expertise review applies the same governance dimensions to documented evidence, interviews, system materials, and relevant policies, producing a written assessment of substantiated strengths, gaps, and remediation priorities.
Want to keep a copy of your results? Enter your email and we’ll send you a summary — no obligation.
Human Continuity Institute — professional ethical review of AI-related documents, policies, and projects, informed by the EU AI Act, OECD AI Principles, and UNESCO’s Recommendation on the Ethics of AI, alongside HCI’s own public-administration-grounded methodology. This checklist is a self-assessment tool and does not constitute a certification, legal conformity assessment, or substitute for independent expert review.
