Red Vault: Security & Sovereignty
Manifest

Unlike traditional Data Loss Prevention (DLP) or Privacy-as-a-Service vendors, Red Vault is built on a Stateless Architecture.

• Zero Data Retention: We do not maintain databases of your PII.
• Volatile RAM Processing: All identification and substitution happens in the browser's active memory. Once the tab is closed or the prompt is sent, the sensitive data is purged from the local environment.
• No Cloud "Phone Home": Your sensitive data is never transmitted to Red Vault servers for "analysis" or "training."

We have replaced "Probabilistic AI" (which guesses what PII is) with a DLP Powered Engine.

• 100% Pattern Match: If a string matches a technical identifier (API Key, SWIFT Code, IP Address), it is caught. Period.
• Zero Hallucination Risk: Because we don't use a secondary LLM to scan your data, there is no risk of the "Privacy AI" leaking your data to the "Public AI."
• Custom Pattern Injection: Enterprise clients can inject proprietary regex patterns for internal project codes, ensuring custom-built protection.

Red Vault utilizes a "Local Secret" architecture to ensure irreversible anonymization.

• Client-Side Salting: Your data is hashed using a cryptographic salt that is stored exclusively within your enterprise infrastructure.
• Irreversible Twins: Even if a third party intercepted the "Synthetic Twin" data, they would require your local hardware-bound salt to reverse the process.
• No Vendor Access: Red Vault employees, engineers, and support staff have mathematically zero access to your original data or your encryption keys.

Our security controls are mapped specifically to the newest Canadian regulatory frameworks:

• Anonymization Standards: Our substitution methodology meets the "Irreversibility" standards required by the Commission d'accès à l'information (CAI).
• Right to Erasure (Law 25): Because Red Vault never stores data, "Right to Erasure" is satisfied by default—the data was never held by the service provider.
• Transparency & Audit: We provide local, exportable audit logs that allow your DPO to verify that 100% of PII was masked before it crossed the border to US-based AI servers.

• Local Sandboxing: The Red Vault extension is sandboxed within the browser, restricted only to the AI domains you white-list.
• SOC2 Type II (In Progress): We are currently undergoing independent auditing to verify our operational security controls.
• Open Schema: We invite client security teams to review our detection logic—transparency is the foundation of trust.