AI Safety, Security, and Governance I
Explore how to implement AI safety, security, and governance controls for generative AI solutions on AWS. Understand strategies to enforce policy compliance, protect sensitive data, minimize hallucinations, and manage secure access using Amazon Bedrock, IAM policies, and data lifecycle features to meet business and regulatory requirements.
We'll cover the following...
Question 38
A company is deploying a customer service chatbot using Amazon Bedrock with Claude 3.5 Sonnet. The chatbot is used by external customers and must comply with strict business policies. One requirement states that the chatbot must never discuss, compare, or reference competitor products or services, even if users explicitly request such information. The company wants a solution that enforces this restriction consistently and requires minimal custom application logic.
Which configuration most directly enforces this requirement with the least custom code?
A. Add system-prompt instructions telling the model to avoid competitor discussions.
B. Configure Amazon Bedrock Guardrails with denied topics for competitor-related content. ...