The integrated endpoint may expose applicant results: What specific risk remains?
Model tests do not establish that the newly integrated endpoint restricts applicant results appropriately.
The question
An employment-screening model passed security testing before integration with a recruiting platform. During integration, the platform adds a shared results endpoint accessible to several internal applications. The team reruns model accuracy tests and scans the model package for malware. What specific risk remains?
Preparing for AIGP? Take the free 5-min readiness quiz →
- The model may produce less accurate rankings after recruiting volume increases.Volume-related performance changes warrant monitoring, but the stated integration change creates a more direct access-control and disclosure risk.
- The original security evidence may become outdated when the supplier releases another version.Future supplier changes require monitoring, but the current unanswered risk arises from the platform endpoint already added during integration.
- The recruiting team may misunderstand the model’s documented confidence scores.User interpretation matters, yet repeated accuracy and package security tests leave the newly introduced endpoint authorization unexamined.
- The integrated endpoint may expose applicant results beyond authorized users. ✓Model-package tests do not evaluate runtime access controls, so integration may create unauthorized disclosure of sensitive applicant information.
The trap
Evaluate newly created interfaces and data paths; pre-integration model testing cannot cover them automatically. How to remember it
Model tests do not establish that the newly integrated endpoint restricts applicant results appropriately.
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
- The supplier update may perform differently: What remaining risk is most specific? →
- The team may miss recurring failure patterns because: What mitigation gap remains? →
- Retraining may reinforce under-investigation because: What risk arises specifically from the label-generation →
- 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.
