NCA-GENL Trustworthy AI Practice Question
An AI governance team is preparing an NVIDIA-hosted LLM for a regulated financial service. They need a documented, repeatable method to detect whether the model produces systematically different approval recommendations for otherwise identical applicants across demographic groups. Which practice best meets this need?
⚠ Common exam trap
The trap here is trusting the model's own statement about its fairness, when self-assessment cannot measure the statistical disparities that counterfactual testing exposes.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Conduct a structured bias evaluation using counterfactual test cases that vary only protected attributes and compare approval rates across groups.
Counterfactual testing is the standard way to isolate disparate treatment: by changing only the protected attribute across otherwise identical applicants, any shift in recommendation is attributable to that attribute, and group-level approval rates quantify the disparity. Leaderboards, self-reports, and latency monitoring cannot produce a reproducible, documented fairness metric tied to matched inputs, so they fail the governance requirement.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Ask the model to self-report whether its own recommendations are biased and record the responses.
Why it's wrong here
Self-reporting relies on the model's introspection, which is unreliable because models lack direct access to the statistical patterns driving their outputs and often produce reassuring but unfounded claims. It also cannot produce a quantitative disparity measure. Recording these statements would document the model's opinion, not evidence of differential treatment across matched applicants.
- ✗
Run the model through a public leaderboard benchmark and publish the aggregate accuracy score.
Why it's wrong here
Leaderboard benchmarks measure general capability across standardized tasks and report a single aggregate number. They do not construct matched applicant pairs or disaggregate outcomes by demographic group, so a high score can coexist with disparate treatment. Aggregate accuracy hides exactly the subgroup differences the governance team must document, making this unsuitable for a fairness audit.
- ✗
Monitor production traffic for anomalous latency spikes that might indicate unequal treatment of certain requests.
Why it's wrong here
Latency measures serving performance and is unrelated to whether recommendations differ by demographic group. A model can treat groups unequally while returning responses in the same time, so latency anomalies would neither reveal nor rule out bias. This conflates operational health monitoring with fairness evaluation and produces no disparity documentation for regulators.
- ✓
Conduct a structured bias evaluation using counterfactual test cases that vary only protected attributes and compare approval rates across groups.
Why this is correct
Counterfactual testing holds all applicant features constant except the protected attribute, so any change in the approval recommendation is attributable to that attribute rather than legitimate risk factors. Comparing approval rates across groups yields a documented, repeatable disparity metric that auditors can reproduce, which is precisely what the governance team needs to evidence systematic differences.
About these practice questions
Courseiva writes every NCA-GENL question from scratch — 367 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official NVIDIA exam blueprint
This NCA-GENL practice question is part of Courseiva's free NVIDIA certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the NCA-GENL exam.