AIncorrect
Fairness
Fairness is the characteristic of preventing specific groups from suffering unfair disadvantage.
This question is about preventing harmful output, not fair treatment between groups, so this is incorrect.
A company is launching a generative AI chatbot for the general public and places top priority on not producing dangerous advice or harmful content. Which responsible AI characteristic BEST matches this idea?
1 / 1
Select an answer
CorrectC
Explanation
A question that asks which responsible AI characteristic prevents harmful output.
- 1「not producing dangerous advice or harmful content」Suppressing harmful output = safety
BIncorrect
Transparency
Transparency is the characteristic of disclosing a model's training data, intended use, and limitations.
This question is about preventing harmful output, not disclosing information, so this is incorrect.
CCorrect
Safety
Correct. Safety is the responsible AI characteristic of keeping AI from producing dangerous or harmful output. Suppressing harmful content with guardrails and similar measures applies here.
DIncorrect
Robustness
Robustness is the characteristic of operating stably without losing performance against input variation or attacks.
It is easily confused with safety, but the core of robustness is that performance does not break down, while the core of safety is that no harmful output is produced. This question is about the latter, so this is incorrect.
Key Takeaway
Not producing dangerous or harmful output is safety, where suppression of harmful content with guardrails and similar measures applies. The easily confused robustness is about performance not breaking down under input variation or attacks; the thing protected differs (robustness = performance, safety = harm in output). Also distinguish it from fairness (between groups) and transparency (disclosure).