QuestionQ66

VerifiedAI Safety, Security, and Governance

A company uses Amazon Bedrock to build a customer-support AI assistant. The AI assistant must answer customer questions about their accounts. It must not disclose personal information in its responses. The company must comply with data-residency policies by ensuring all processing takes place in the same AWS Region in which each customer is located.

Before making the AI assistant available to customers, the company wants to assess how effectively it prevents personal-information exposure.

Which solution meets these requirements?

Explanation

Amazon Bedrock Guardrails sensitive-information filters are intended to block or mask personally identifiable information in model inputs and responses. Masking during development and testing supports safe assessment of the filtering behavior, while blocking in production prevents detected personal information from being returned. Guardrail evaluation must remain in the customer's Region because cross-Region guardrail inference can route input prompts and output results outside the primary Region; deploying an equivalent guardrail in each operating Region preserves the single-Region processing requirement.

Learn more

Community Discussion

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