AI Governance

AI governance is being developed
as part of API Platform.

We are developing a mechanism to manage rules scattered across AI tools as part of API Platform. We are validating RBAC and audit log in Private Beta; other items are future considerations.

Why

As more tools and teams adopt AI, usage rules, data terms, and system instructions are managed individually. We are developing a unified management foundation so consistent rules and audits can apply company-wide.

Rule hierarchy
  1. Enterprise policy
  2. Department policy
  3. Project policy
  4. System instructions
  5. Exceptions
  6. Approval
  7. Audit

Validatable in Private Beta

RBACPrivate Beta
Audit logPrivate Beta

Future considerations

Policy RegistryConcept
Approval flowConcept
VersioningConcept
Drift DetectionConcept

Approval, publication, exception management, and audit are performed based on the authority of people and the organization. AI can help draft and improve policy, but the final decision and responsibility rest with people.