A company launching a customer-facing chatbot needs configurable filters that block harmful content, redact PII, and restrict disallowed topics — applied consistently across whichever foundation model they use. What's the AWS-native control?
A.Amazon Bedrock Guardrails
Correct. Guardrails is the purpose-built safety layer for generative AI apps.
B.AWS WAF rules
WAF filters web traffic patterns, not model output content.
C.Amazon Macie findings
Macie discovers sensitive data in S3 — it doesn't filter model responses.
D.CloudTrail data events
CloudTrail records API calls — no content filtering.
Why this answer
Bedrock Guardrails provide model-agnostic safeguards: content filters by harm category, denied topics, PII redaction, and word filters — applied at inference regardless of the underlying FM.
Correct answer: A · Last reviewed 2026-10-26
Official AWS references
Practice this interactively
Answer without seeing the solution first in the Responsible AI practice session.
Practice Responsible AI