AI Governance Framework
A structured approach to governing AI systems — typically a document that outlines principles, processes, roles, and controls for AI governance.
An AI governance framework is the structured approach an organisation takes to governing its AI systems. Major frameworks include:
- NIST AI Risk Management Framework (US) - EU AI Act (Europe) - ISO/IEC 42001 (International) - Singapore AI Governance Framework - OECD AI Principles
These frameworks typically address: - Risk categorisation (what level of risk does each AI system present?) - Accountability (who is responsible for each AI system?) - Transparency (how are AI decisions explained?) - Bias and fairness (how is discrimination prevented?) - Security and privacy (how is data protected?)
The limitation of frameworks is that they are documents, not infrastructure. A framework describes what governance should look like; it doesn't enforce it. This is the gap between governance frameworks and governance infrastructure.
How Constellation handles this
Constellation implements the infrastructure that governance frameworks describe. Where a framework says "organisations should enforce constraints on AI systems," Constellation provides the governance gate that actually enforces those constraints in real time.