Governance Guide

Governing AI Development Through Enterprise Architecture

Align AI-assisted delivery with architecture, security, data governance, testing, accountability, and procurement controls.

Govern the outcome

AI does not change accountability

The team and organization remain responsible for generated code, data use, security, licensing, testing, operations, and architectural alignment.

Required controls

Place AI inside the existing delivery lifecycle

  • Approved tools and data-classification rules
  • No secrets or restricted data in unapproved models
  • Architecture decision and interface conformance
  • Human code review and security testing
  • Dependency and licence review
  • Traceable prompts and generated artifacts where required
  • Production observability and rollback

Enterprise boundary

AI-created applications consume the same governed shared services

Authentication, application trust, notifications, payments, integration, and support remain platform responsibilities regardless of how application code was produced.

Start the conversation

Continue the architecture conversation

Explore how these principles apply to your organization’s current systems and transformation goals.