Also known as: AI Security Governance
The policies and technical controls that define which AI systems may use which data, perform which actions, and produce which evidence under human oversight.
AI governance defines the allowed use of models, data, tools, workflows, and automated actions. In a SIEM, that includes model selection, data handling, prompts, tenant boundaries, roles, approvals, action limits, audit history, and review of the evidence behind a result.
Governance should become stricter as the system moves from summarizing information to changing security or business systems.
AI governance is not only a policy document. The important controls have to be enforced in permissions, workflow logic, logs, and approval steps.