CASE 04 / Operational Compliance
If the system makes the wrong call, who answers?
An AI inventory, written policy and named oversight make accountability visible before a consequential decision goes wrong.
UNLIT-reported case · Source notes belowTo a documented AI use policy
High-risk systems assessed in the engagement
Written operating rules and oversight
Post-engagement AI governance audit findings
01 / THE CHALLENGE
The problem behind the problem.
The case-study collection describes organizations using AI across operational decisions without a shared inventory, written policy or clear accountability chain. Leadership could not readily identify who owned a consequential error.
The work organized the uses, risks and responsibilities into a reviewable governance package. The value was a decision structure that people could follow and leadership could explain.
02 / THE APPROACH
From diagnosis to daily practice.
- 01
Inventory the systems
Document AI tools used in procurement, scheduling, document processing and other workflows, including uses outside the formal rollout.
- 02
Assess consequential uses
Review bias risk, privacy, decision accountability and potential impact on affected people.
- 03
Write the operating policy
Define approved and prohibited uses, human oversight, training expectations, escalation and review responsibilities.
- 04
Make it board-ready
Prepare an executive summary that identifies decisions, owners, remaining risks and review needs.
THE TAKEAWAY
Assign decision ownership and a path for human review before relying on an automated recommendation.
YOUR TURN / A PRACTICE SCENARIO
What would you do?
An AI recommendation affects someone’s access to a service. No one can name the accountable decision-maker. What comes first?
An educational scenario developed from the case theme; your selection is not a personality score or a finding about the original engagement.
FOR YOUR NEXT INTERNAL CONVERSATION
A scope you can discuss.
Use these outputs to frame a needs assessment, an internal proposal or a conversation with UNLIT. Define acceptance criteria, evidence and responsibilities before an engagement begins.
- AI system inventory
- Risk and impact assessment
- Written AI use policy
- Executive governance summary
TAKE THE NEXT DEGREE WITH YOU
Your decision. Your next move.
Use QUAD (Question, Understand, Analyze, Decide) to turn the idea into an action you can review.
Your answers stay in this browser unless you download or share them. Saved plans do not sync between devices. Avoid confidential information on a shared device.