The Claude Compliance Collector is a customer-native deployed CloudFormation stack that automatically retrieves Claude Enterprise conversation data using the Anthropic Compliance API. The conversation data is scanned for policy violations, which then appear as detections in TrendAI Vision One™ AI Guard.
The Claude Compliance Collector runs on a schedule in your AWS environment, pulling
conversation content directly from the Anthropic Compliance API and scanning it for
policy violations using AI Guard. You can review and investigate detections in the
AI Application Security dashboard, giving you the visibility and audit trail that
compliance programs require.
Your Anthropic Compliance Access Key never leaves your infrastructure. Conversation
content is forwarded to AI Guard for scanning but is not stored outside your environment.
How It Works
After deployment, the collector runs automatically on a configurable schedule (every
four hours by default) and completes the following steps:
-
Scans the Anthropic Compliance API Activity Feed to discover active Claude chat IDs within the configured time window, automatically excluding deleted chats.
-
Retrieves conversation content for each discovered chat, including prompts, model responses, and any text content sent alongside uploaded files.
-
Chunks large transcripts and forwards them to AI Guard for policy violation detection.
-
AI Guard detections appear in the TrendAI Vision One™ AI Application Security dashboard, where you can filter results by the collector's application name and investigate flagged interactions. For more information about AI Guard, see AI Application Security.
AWS Resources
Deploying the collector creates the following resources in your AWS account:
-
Lambda functions
-
SQS queue
-
EventBridge schedule
-
Secrets Manager secrets (stores API keys encrypted at rest)
-
IAM roles
-
CloudWatch Log Groups
Requirements
Before deploying the collector, verify that your environment meets the prerequisites.
See Claude Compliance Collector prerequisites.