A company is using a foundation model via Amazon Bedrock to generate legal documents. They want to ensure the model's outputs are transparent and that users can understand the basis for the generated content. Which combination of actions would best achieve this?

Use a retrieval-augmented generation (RAG) approach that cites sources, and provide a confidence score.

Use a fine-tuned model and do not provide any additional information.

Use a human review process for all generated documents.

Use Amazon Bedrock Model Evaluations to assess the model's accuracy and provide a model card.

A healthcare startup is using a deep learning model to predict patient readmission risk. To meet internal governance requirements, they need to document the model's intended use, training data, evaluation metrics, and known limitations in a standardized format that can be shared with clinicians. Which AWS service should they use?

Amazon Bedrock Model Evaluations

An autonomous vehicle company uses a deep neural network for object detection. They want to add a post-hoc explainability tool to generate saliency maps for each detection. However, the explainability tool requires a simplified surrogate model that reduces the system's ability to detect rare edge cases. What is the most appropriate course of action?

Deploy the surrogate model to ensure full explainability, even if it reduces safety.

Use the surrogate model only for non-critical features like lane detection.

Abandon the use of AI in the vehicle to avoid the tradeoff.

Keep the original model and use explainability tools only for post-incident analysis.

A machine learning engineer is evaluating two models for a credit scoring system. Model X is a linear model with 10 features. Model Y is a deep neural network with 1000 features. Both have similar accuracy. Which model is more likely to be considered transparent, and why?

Model X, because its predictions can be understood by examining the feature weights.

Both models are equally transparent.

Model Y, because it has more features.

What is a key principle of human-centered design for explainable AI?

Explanations should only be provided to technical experts.

Explanations should be generated automatically without user feedback.

Explanations should be tailored to the needs and expertise of the end user.

Explanations should always be as detailed as possible, regardless of the audience.

AWSCertified AI Practitioner

Domain 4Objective 2

Task Statement 4.2: Recognize the Importance of Transparent and Explainable Models. AIF-C01 Practice Questions (Page 4)

Part of the Content Domain 4: Guidelines for Responsible AI domain, which makes up ~11% of our current practice bank. AWS does not publish an official question count, but from its 90-minute exam (~35–60 total, ~4–7 in this domain), expect 2–4 from this objective — we provide 21 practice questions to prepare you well beyond it. (estimate)

21questions here

5free pages

4concepts

Questions 16–20

16expert · hard
A company is using a foundation model via Amazon Bedrock to generate legal documents. They want to ensure the model's outputs are transparent and that users can understand the basis for the generated content. Which combination of actions would best achieve this?
Select an answer first
17application · medium
A healthcare startup is using a deep learning model to predict patient readmission risk. To meet internal governance requirements, they need to document the model's intended use, training data, evaluation metrics, and known limitations in a standardized format that can be shared with clinicians. Which AWS service should they use?
Select an answer first
18expert · hard
An autonomous vehicle company uses a deep neural network for object detection. They want to add a post-hoc explainability tool to generate saliency maps for each detection. However, the explainability tool requires a simplified surrogate model that reduces the system's ability to detect rare edge cases. What is the most appropriate course of action?
Select an answer first
19application · medium
A machine learning engineer is evaluating two models for a credit scoring system. Model X is a linear model with 10 features. Model Y is a deep neural network with 1000 features. Both have similar accuracy. Which model is more likely to be considered transparent, and why?
Select an answer first
20foundation · easy
What is a key principle of human-centered design for explainable AI?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to AIF-C01

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by AWS. “AIF-C01” is a trademark of its owner, used for identification only.

Task Statement 4.2: Recognize the Importance of Transparent and Explainable Models. AIF-C01 Practice Questions (Page 4)

Questions 16–20

A company is using a foundation model via Amazon Bedrock to generate legal documents. They want to ensure the model's outputs are transparent and that users can understand the basis for the generated content. Which combination of actions would best achieve this?

A machine learning engineer is evaluating two models for a credit scoring system. Model X is a linear model with 10 features. Model Y is a deep neural network with 1000 features. Both have similar accuracy. Which model is more likely to be considered transparent, and why?

What is a key principle of human-centered design for explainable AI?

Keep studying — Content Domain 4: Guidelines for Responsible AI