Exams / AIF-C01 / Responsible AI and guardrails / Question 1
AIF-C01 practice question: Responsible AI and guardrails (1 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.
A bank's customer chatbot must never give personal investment advice, even if a customer asks for it directly. Which Amazon Bedrock Guardrails feature is designed to block this?
Choose one answer.
- AA. Contextual grounding checks
- BB. Word filters
- CC. Sensitive information filters
- DD. Denied topics
Show the answer and explanation
Answer: D
- A. Wrong: Grounding checks detect responses not supported by the source material. They do not block a whole subject.
- B. Wrong: Word filters block specific words or phrases. Investment advice can be phrased in endless ways, so a word list would miss much of it.
- C. Wrong: These detect and mask data such as PII. They do not block a type of advice.
- D. Right: Correct. Denied topics let you define subjects the application must not discuss and block them in prompts and responses.
Explanation
Amazon Bedrock Guardrails offers several policies: content filters for harmful categories and prompt attacks, denied topics, word filters, sensitive information filters, contextual grounding checks and Automated Reasoning checks. Denied topics are defined in natural language, so they catch a subject however it is phrased.
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.