AIF-C01 Question 166
Single answerA company is using Amazon Bedrock to build and deploy a generative AI application. They want to ensure that their AI model adheres to responsible AI principles, such as preventing harmful or biased outputs. Which feature of Amazon Bedrock can help them implement guardrails for responsible AI?
- A
Customizing model outputs using prompt templates
- B
Utilizing Guardrails to enforce content moderation and bias detection
- C
Deploying models with Amazon SageMaker for end-to-end monitoring
- D
Using AWS Key Management Service (KMS) to encrypt sensitive data
Show answer and explanation
Correct answer: B
Explanation
Guardrails in Amazon Bedrock are specifically designed to help developers implement responsible AI practices, such as preventing harmful or biased outputs. This ensures that applications built using generative AI models operate within ethical and legal boundaries.
- A. Incorrect.
While prompt templates can help structure model responses, they do not directly address responsible AI principles like content moderation or bias detection.
- B. Correct.
Amazon Bedrock's Guardrails feature is designed to enforce responsible AI principles, such as content moderation and bias detection, making it the correct choice.
- C. Incorrect.
Amazon SageMaker provides monitoring capabilities for ML models but is not specific to implementing guardrails for responsible AI in Amazon Bedrock.
- D. Incorrect.
AWS Key Management Service (KMS) is used for encrypting sensitive data but does not contribute to enforcing responsible AI principles in model outputs.