How AI Governance Check works
AI Governance Check is a free decision-support tool that helps small and mid-sized organizations evaluate a proposed or existing AI use and determine how much governance it may require. It is a companion to the Practical AI Governance Toolkit from LTECwithLance.
You first describe the intended use in your own words. The AI context analyst extracts relevant context such as affected people, higher-impact domains, external exposure, vendor involvement, and autonomy. You review and correct that interpretation before it is used. You then answer four short structured questions about information sensitivity, AI authority and human review, the consequences of error, and whether the intended purpose may involve prohibited or clearly unacceptable conduct. Moderate, high-impact, and stop / escalate uses receive one additional safeguards question.
The final classification is produced by a deterministic rules engine, not by the AI analyst. The analyst supplies context, while the user confirms material facts and the rules engine calculates inherent risk, governance readiness, and any specialist-review routing. This keeps the result explainable and auditable.
The design follows a simple principle: complexity should increase only when risk increases. Routine drafting or summarization should not require the same governance process as employment screening, clinical decision support, lending, biometrics, or autonomous actions with real-world consequences.
The five governance stages
Know
Identify which AI systems and use cases exist across the organization. You cannot govern what you cannot see, so start with a clear inventory of AI tools, vendors, and intended purposes.
Own
Assign a named business owner and accountability for each material AI use. Clear ownership ensures someone is responsible for outcomes, risks, and ongoing review.
Assess
Determine the level and nature of risk. Consider data sensitivity, impact on people, system authority, high-impact domains, and the consequences of error.
Control
Apply safeguards appropriate to the assessed risk. Higher-impact uses require stronger controls such as human oversight, approval, testing, transparency, and monitoring.
Monitor
Review the system over time and respond to problems. AI systems, vendors, and operating conditions change, so governance is an ongoing practice.
Ready to evaluate an AI use?
Start assessment