QuestionQ96

VerifiedAI Safety, Security, and Governance

A company uses Amazon Bedrock to build an application that helps researchers apply for grants. The application is based on an Amazon Nova Pro foundation model (FM). It has four required inputs and must return responses in a consistent text format. The company wants to receive a notification in Amazon Bedrock when a response contains bullying language, but does not want every flagged response to be blocked.

The company creates an Amazon Bedrock flow that accepts an input prompt and sends it to the Amazon Nova Pro FM. The Amazon Nova Pro FM returns a response.

Which additional steps must the company take to meet these requirements?

Choose two
Explanation

Amazon Bedrock Prompt Management supports prompt variables, model selection, and instructions that specify an expected response format; a flow prompt node can use such a managed prompt. Amazon Bedrock Guardrails defines bullying as the Insults content category. Its Detect action returns detection information without blocking the model response, whereas Block replaces the content with blocking messaging.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!