A team must block a chatbot from discussing competitors and remove personal data from replies. What should they use?
Why A is the answer
Amazon Bedrock Guardrails
Guardrails handle denied topics and sensitive information filtering in one place.
In detail
- Guardrails can block topics and remove personal data.
- Prompt instructions alone are easier to bypass.
Use case
An illustrative scenario, not a real incident.
A retailer's chatbot must not mention rival brands or repeat phone numbers. Guardrails enforce both, outside the prompt.
See it in different ways
flowchart TD
A["User input"] --> B{"Bedrock Guardrails"}
B -->|"Injection, jailbreak or blocked topic"| C["Block"]
B -->|"Allowed"| D["Foundation model"]
D --> E{"Check the output"}
E --> F["Response"]mindmap
root((Responsible AI))
Bias
SageMaker Clarify
Documentation
SageMaker Model Cards
Safety
Bedrock Guardrails
Transparency
User feedbacksequenceDiagram participant P0 as User participant P1 as Guardrail participant P2 as Model P0->>P1: Input P1->>P2: Allowed input P2-->>P1: Output P1-->>P0: Filtered output
Key point: Guardrails enforce content rules.
Why the other options are wrong
- B. A bigger foundation model
- A bigger model adds no controls.
- C. Amazon S3 versioning
- Versioning protects objects.
- D. AWS Cost Explorer
- Cost Explorer shows spending.