bedrock:CreateGuardrail

Write

bedrock:CreateGuardrail is a Write access-level action in the bedrock service. Grants permission to create a new guardrail

Resource types

ARNs that bedrock:CreateGuardrail can be scoped to in an IAM policy.

arn:${Partition}:bedrock:${Region}:${Account}:automated-reasoning-policy/${AutomatedReasoningPolicyId}

automated-reasoning-policy

arn:${Partition}:bedrock:${Region}:${Account}:automated-reasoning-policy/${AutomatedReasoningPolicyId}:${AutomatedReasoningPolicyVersion}

automated-reasoning-policy-version

arn:${Partition}:bedrock:${Region}:${Account}:guardrail-profile/${ResourceId}

guardrail-profile

Condition keys

Keys you can use in the Condition block of a policy that grants bedrock:CreateGuardrail.

aws:ResourceTag/${TagKey} String

Filters access by having actions based on the tag value associated with the resource

Example IAM policy

A minimal policy that allows bedrock:CreateGuardrail. Replace "Resource": "*" with a specific ARN to follow least privilege.

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Sid": "AllowBedrockCreateGuardrail",
      "Effect": "Allow",
      "Action": "bedrock:CreateGuardrail",
      "Resource": "*"
    }
  ]
}

Related actions in bedrock

bedrock:AgenticRetrieveStream Read bedrock:AllowVendedLogDeliveryForResource Permissions management bedrock:ApplyGuardrail Read bedrock:AssociateAgentCollaborator Write bedrock:AssociateAgentKnowledgeBase Write bedrock:AssociateThirdPartyKnowledgeBase Write bedrock:BatchDeleteAdvancedPromptOptimizationJob Write bedrock:BatchDeleteEvaluationJob Write bedrock:CallWithBearerToken Read bedrock:CancelAutomatedReasoningPolicyBuildWorkflow Write bedrock:CopyBlueprintStage Write bedrock:CountTokens Read

View all bedrock actions →