ACorrect
Filtering of harmful content such as hate and violence
Correct. Guardrails can configure a protection that detects and filters harmful content such as hate and violence.
A company is checking how far Bedrock Guardrails can meet the safety requirements for a chatbot. Which protections (filters) can be configured with Amazon Bedrock Guardrails? (Choose TWO.)
1 / 1
Select all that apply
CorrectA, B
Explanation
A question that asks to choose TWO protections that can be configured with Guardrails.
- 1「protections (filters) can be configured」The filtering features of Guardrails
- 2「Amazon Bedrock Guardrails」Harmful content and PII filtering apply
BCorrect
Detection and masking of personal information
Correct. Guardrails can configure a protection that detects personal information (PII) such as names, addresses, and card numbers, and masks or blocks it.
CIncorrect
Access permission control with IAM
Access permission control is the role of IAM and is not a Guardrails configuration item.
Guardrails handles the content of the model input and output, so it is incorrect.
DIncorrect
Metric monitoring of the number of model invocations
Metric monitoring is the role of CloudWatch and is not a Guardrails feature.
Guardrails handles content filtering, so it is incorrect.
EIncorrect
Management of the vector search index
Vector index management is a component on the knowledge base (RAG) side.
It is not a Guardrails protection feature, so it is incorrect.
Key Takeaway
The protections that can be configured with Bedrock Guardrails include 'filtering harmful content (hate, violence, and so on)' and 'detecting and masking personal information (PII)' (along with blocking denied topics, blocking prompt attacks, and automated reasoning checks). Locking the server rack (physical security), air-conditioning temperature, and logo resolution are none of them Guardrails output filter configuration items.