Exams / AIF-C01 / Responsible AI and guardrails / Question 2
AIF-C01 practice question: Responsible AI and guardrails (2 of 5)
Guidelines for Responsible AI A free question from our original AIF-C01 practice exam. Try it, then open the answer to see why each option is right or wrong.
Before launch, a utility company tests its billing assistant with misspelt questions, unusual phrasing and deliberate attempts to confuse it, to check that it still gives correct answers. Which dimension of responsible AI is the team testing?
Choose one answer.
- AA. Veracity and robustness
- BB. Fairness
- CC. Transparency
- DD. Governance
Show the answer and explanation
Answer: A
- A. Right: Correct. This dimension is about correct outputs, even with unexpected or adversarial inputs.
- B. Wrong: Fairness considers impacts on different groups of people. The tests here are about input quality, not groups.
- C. Wrong: Transparency helps people make informed choices about using an AI system. It is not about handling odd inputs.
- D. Wrong: Governance brings best practices into the AI supply chain. It is not tested by sending unusual inputs.
Explanation
AWS describes responsible AI through dimensions including fairness, explainability, privacy and security, safety, controllability, veracity and robustness, governance and transparency. Testing with noisy, unexpected and adversarial inputs checks veracity and robustness.
Study
Keep practising
Answer questions with a score at the end, or practise just this topic with the explanation after each one. No sign-up needed.