Probabilistic outputs and opacity: Which pair of AI characteristics best explains this governance challenge?
Different outputs for identical inputs reflect probabilistic behavior, and untraceable reasoning reflects opacity.
The question
A hospital's diagnostic model sometimes returns different recommendations for identical patient inputs, and clinicians cannot reconstruct the reasoning behind any single output. Which pair of AI characteristics best explains this governance challenge?
Preparing for AIGP? Take the free 5-min readiness quiz →
- Autonomy and speed, because the system acts without human input and processes cases faster than staff can reviewPlausible but wrong: autonomy and speed do not explain why identical inputs give different, untraceable outputs.
- Data dependency and complexity, because output quality hinges on training data and the model has many interacting internal partsAlmost right: complexity contributes to opacity, but data dependency does not explain non-repeatable outputs for identical inputs.
- Probabilistic outputs and opacity, because results are non-deterministic and the internal reasoning is not interpretable ✓Correct: varying outputs for identical inputs reflect probabilistic (non-deterministic) behavior, and untraceable reasoning reflects opacity.
- Scale and potential for misuse, because the model is deployed widely and could be repurposed beyond clinical usePlausible but wrong: scale and misuse are real characteristics but are unrelated to output variability and lack of traceability.
The trap
Attributing non-repeatable, unexplainable outputs to autonomy or speed rather than to probabilistic behavior and opacity. How to remember it
Different outputs for identical inputs reflect probabilistic behavior, and untraceable reasoning reflects opacity.
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
- Ethics and bias risk: Which category of AI risk does this situation most directly illustrate? →
- Transparency and explainability: Which responsible-AI principle is the bank primarily operationalizing? →
- An AI governance committee or council: Which structure best fits this responsibility? →
- 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.
