AI Governance glossary
44 terms on governing AI systems, models and agents.
Frequently asked questions
What is the AI Governance glossary?
It is a reference of more than 40 terms used to govern how organizations build, buy and use AI, from AI inventories, risk assessments and model validation to the EU AI Act, ISO 42001 and agentic AI governance. Each entry gives a short definition first, then explains what the term means in practice.
Who is the AI Governance glossary for?
AI governance and responsible AI leads, chief AI officers, risk, compliance and legal teams, data science and machine learning leaders, security teams, and anyone accountable for how AI is approved and used.
How are the terms organized?
Terms are listed A to Z and tagged by category: governance and accountability; AI risk and assurance; regulation and standards; and model and data lifecycle. Use the search box or the category filters to narrow the list, or jump straight to a letter.
Where should AI governance start?
Most organizations start with an AI inventory: one register of every AI system, model and AI-enabled tool in use, with an owner, purpose and risk tier for each. Risk assessments, approvals, monitoring and regulatory reporting all depend on knowing what is running. See AI inventory.
How does AI governance fit into the Enterprise Operating Model?
AI governance sits in the Govern & Deliver part of the Enterprise Operating Model. It keeps one register of the AI an organization runs, connects each system to its owner, risks and controls, and shows what it costs and what it returns. Explore AI governance.