Define permitted-use boundaries with technical access: Which measure most directly addresses the concern?
Scoping permitted use and enforcing it with access and usage controls directly prevents the foreseeable secondary, unintended use.
The question
A firm plans to release an internal document-summarization model to employees, but governance realizes the same capability could be repurposed to profile individuals from combined HR records, a use never intended or reviewed. The team must reduce the risk of this foreseeable secondary use before rollout. Which measure most directly addresses the concern?
Preparing for AIGP? Take the free 5-min readiness quiz →
- Increase the model's summarization accuracy through additional fine-tuning so its condensed outputs represent the underlying source documents more faithfully.Plausible as quality work, but higher summarization accuracy does nothing to prevent the capability from being redirected toward unintended profiling.
- Expand server capacity so the summarization service can scale to every employee without latency during periods of peak concurrent demand.Plausible as deployment planning, but scaling infrastructure addresses availability, not the downstream-harm risk of a foreseeable secondary use.
- Publish a transparency statement telling employees that an AI system generates the document summaries they receive through the tool.Plausible as disclosure, but merely informing users that AI is in use does not restrict or prevent the profiling misuse the scenario forecasts.
- Define permitted-use boundaries with technical access controls and usage restrictions that block the model from being applied to the profiling scenario. ✓Correct because forecasting and reducing secondary or unintended uses is best served by scoping permitted use and enforcing it with controls that foreclose the foreseeable misuse.
The trap
Believing that disclosing AI use or improving output quality reduces the risk of a system being repurposed for unintended harm. How to remember it
Scoping permitted use and enforcing it with access and usage controls directly prevents the foreseeable secondary, unintended use.
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 Deployment and Use questions
- Record the event in an incident log capturing cause: Which documentation action best satisfies this governance →
- To define how the organization informs and responds: What is the primary purpose of such a plan? →
- To be able to shut down or geographically restrict: What is the main governance reason for having such a →
- All 424 Understanding How to Govern AI Deployment and Use questions →
Part of the Certsqill AIGP question bank · Understanding How to Govern AI Deployment and Use ·
Every answer, right and wrong, comes with its own explanation.
