Sovereignty Auditor

The Sovereignty Auditor ensures that your AI model is executing within the required geographical and legal jurisdiction by using cryptographic "Landmark" probes.

🛡️ Use Case

• Data Residency: Enforced compliance with US/EU data sovereignty laws by proving the TEE is hardware-anchored to a specific data center.
• Geofencing: Prevent sensitive model inference in unauthorized regions.

📝 Implementation

This auditor performs an active probe to "Anchor Nodes" in the Request phase.

import os
from lucid_sdk import create_auditor, Proceed, Deny

builder = create_auditor(auditor_id="sovereignty-auditor")
REQUIRED_REGION = os.getenv("REQUIRED_JURISDICTION", "US")

@builder.on_request
def check_sovereignty(data: dict):
    # The SDK's underlying LucidClient communicates with 'Landmark' nodes
    # to verify the hardware's physical proximity and jurisdiction.

    # Simple check for demo purposes:
    current_region = os.getenv("NODE_REGION", "US")

    if current_region != REQUIRED_REGION:
        return Deny(
            reason=f"Jurisdiction violation: Workload in {current_region}, policy requires {REQUIRED_REGION}"
        )

    return Proceed(jurisdiction=current_region)

auditor = builder.build()

☸️ Deployment Configuration

Add this to your auditors.yaml. Notice the use of environment variables to configure the required jurisdiction.

chain:
  - name: sovereignty-auditor
    image: "lucid/sovereignty-auditor:v1"
    port: 8083
    env:
      REQUIRED_JURISDICTION: "US"
      ANCHOR_URLS: "http://anchor-dc1.lucid-system.svc.cluster.local:8001"

🔍 Behavior

• Verification: The auditor queries local Anchor nodes for signed receipts.
• Attestation: These receipts are sent to the Verifier, and the final AI Passport includes a "Location Verified" assertion.