Learning & Improvement closes the GenAI governance loop: feedback from live systems prompts changes to policies, designs, processes, tests, or controls. Product owners and system stewards drive application-level changes, while the Responsible-AI steering committee reviews the overall strategy and key performance indicators. section 1.3 section 3.9
Use signals from operations to guide improvement:
- Feed customer complaints, red-team findings, and new regulations into charter revisions, refreshed policies, or engineering work. section 1.3
- Track metrics across governance levels to check whether controls remain effective and identify where more work is needed. section 3.9
Learning from a live-system signal
For a hiring system, periodically audit whether recommendations produce discriminatory impact; use findings to guide governance or system changes. section 1.3
Figure 1.4 Six Levels of Generative AI Governance. Chapter 2 will expand into what control tasks attach to each checkpoint.

Locate Learning & Improvement as the final checkpoint in the six-level governance model. section 1.3
