Structures, rules and practices that ensure artificial intelligence systems are safe, lawful, fair and accountable.
Also known as: Responsible AI, AI assurance
In plain English
Artificial intelligence governance is like the road rules and safety checks for AI systems. It defines who is responsible, what behaviour is allowed, and what protections must be in place—similar to parental controls and app permissions on your phone.
In practice
At work it means policies, risk assessments, supplier checks and monitoring you apply when buying or deploying AI tools. Good governance reduces legal, reputational and operational risk and helps meet regulators' and customers' expectations.
Under the hood
AI governance combines policy frameworks and lifecycle controls with technical measures such as access control, versioning, data lineage, logging, and metrics for fairness, robustness and privacy. It relies on audits, continuous monitoring, policy-as-code, documentation and traceability to enforce compliance across development, deployment and maintenance.
Example
“We need to update our artificial intelligence governance checklist before approving the vendor's model for production.”