AWS bedrock documentation change
Summary
Added Automated Reasoning checks as a safeguard policy and updated API references
Security assessment
Documents new Automated Reasoning checks to validate model responses, a security feature. Enhances security documentation but does not address a specific vulnerability.
Diff
diff --git a/bedrock/latest/userguide/guardrails.md b/bedrock/latest/userguide/guardrails.md index f21850d22..b82008381 100644 --- a//bedrock/latest/userguide/guardrails.md +++ b//bedrock/latest/userguide/guardrails.md @@ -31,0 +32,2 @@ Amazon Bedrock Guardrails provides the following safeguards (also known as polic + * **Automated Reasoning checks** – Can help you validate the accuracy of foundation model responses against a set of logical rules. You can use Automated Reasoning checks to detect hallucinations, suggest corrections, and highlight unstated assumptions in model responses. + @@ -39 +41 @@ Experiment and benchmark with different configurations and use the built-in test -Guardrails can be used directly with FMs during the inference API invocation by specifying the guardrail ID and the version. Guardrails can also be used directly through the ApplyGuardrail API without invoking the foundation models. If a guardrail is used, it will evaluate the input prompts and the FM completions against the defined policies. +Guardrails can be used directly with FMs during the inference API invocation by specifying the guardrail ID and the version. Guardrails can also be used directly through the `ApplyGuardrail` API without invoking the foundation models. If a guardrail is used, it will evaluate the input prompts and the FM completions against the defined policies.