Provide reasons: Which control is most defensible?
Meaningful oversight requires information, authority, and escalation targeted to foreseeable harm rather than merely inserting a human.
The question
A customer-service model recommends denying refund requests. Agents currently receive only the recommendation and a confidence score; some cases require rapid handling, while disputed cases may cause financial harm. The organization wants meaningful human oversight without making every decision manual. Which control is most defensible?
Preparing for AIGP? Take the free 5-min readiness quiz →
- Require agents to review every recommendation manually before responding.Universal manual review supplies human involvement but does not ensure informed intervention without reasons, authority, or escalation design.
- Provide reasons, authority to override, and escalation for disputed or harmful cases. ✓This combines usable information, intervention authority, and proportionate escalation while preserving efficient handling of routine cases.
- Route all low-confidence recommendations to supervisors and retain the current agent workflow otherwise.Confidence routing may miss harmful high-confidence errors and still leaves agents without sufficient information or explicit override authority.
- Let agents accept recommendations unless customers appeal.This leaves agents without reasons or proactive escalation, making nominal review vulnerable to automation bias and delayed correction.
The trap
Check information, time, authority, override, fallback, and escalation—not merely whether a person appears in the workflow. How to remember it
Meaningful oversight requires information, authority, and escalation targeted to foreseeable harm rather than merely inserting a human.
How many of these would you get right?
One of 1581 AIGP questions on Certsqill. Take a free five-minute check and see your score per domain — not one number, but which section to open tonight.
Test your AIGP readiness — freeMore Understanding the Foundations of AI Governance questions
- Assign university owners to validate context: Which responsibility should the university retain? →
- State thresholds: Which communication best conveys the organization’s risk tolerance? →
- Retain internal accountability and monitor vendor: Which arrangement best reflects the retailer’s governance →
- All 337 Understanding the Foundations of AI Governance questions →
Part of the Certsqill AIGP question bank · Understanding the Foundations of AI Governance ·
Every answer, right and wrong, comes with its own explanation.
