Low-confidence itinerary answers may still be inaccurate: Before accepting residual risk, which demonstrated
Controls reduce itinerary errors but do not eliminate inaccurate answers issued before escalation.
The question
A travel-support assistant is authorized for itinerary questions only. It applies confidence thresholds and escalates low-confidence itinerary answers. Testing shows fewer booking errors, but users continue asking visa questions; the assistant is configured to refuse those questions and route users to official guidance. Before accepting residual risk, which demonstrated risk must be documented?
Preparing for AIGP? Take the free 5-min readiness quiz →
- Low-confidence itinerary answers may still be inaccurate before escalation. ✓Thresholds and escalation reduce exposure but do not guarantee that an answer issued before intervention is accurate.
- Users may mistake itinerary guidance for confirmed booking changes.That interpretation risk would require evidence that users treat guidance as booking confirmation; it is not the demonstrated residual risk.
- The assistant may exceed its itinerary scope.The scenario states that visa questions are refused and routed, so scope violation is not the demonstrated residual risk.
- Peak-period escalation queues may become unavailable.Capacity could be an operational risk, but the scenario provides no evidence that queues are unavailable or overloaded.
The trap
A control that reduces risk does not necessarily eliminate the event being controlled. How to remember it
Controls reduce itinerary errors but do not eliminate inaccurate answers issued before escalation.
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 How to Govern AI Development questions
- Retain treatment and monitor severe misuse pathways: Which governance conclusion is most defensible? →
- Unmeasured subgroup false negatives may delay or deny: Which remaining risk most directly concerns →
- High-confidence routing errors can bypass the human: Which remaining risk should be addressed before →
- All 426 Understanding How to Govern AI Development questions →
Part of the Certsqill AIGP question bank · Understanding How to Govern AI Development ·
Every answer, right and wrong, comes with its own explanation.
