01. Private Access Model
Safepad AI is a restricted-access security tool. It does not allow public registration. Credentials are generated in-house and issued exclusively to existing Digital Safety clients. We do not store, request, or collect user names, phone numbers, or social identities.
02. Hardware Permissions
To provide AI safety features, Safepad AI requires access to the Microphone and Camera.
- Microphone: Used for real-time Speech-to-Text. Initial processing is performed on-device.
- Camera: Used for visual AI analysis to provide security insights.
- Encryption: All sensor data transmitted for advanced AI analysis is encrypted in transit using industry-standard protocols.
03. Data Sovereignty
Safepad AI maintains strict data residency requirements:
- Location: All cloud-synced data is stored exclusively in Stockholm, Sweden.
- Persistence: AI-analyzed results and safety logs are stored in MongoDB and Vercel Blob.
- Offline Mode: If used without a Digital Safety login, all data is stored strictly on the device’s local storage and is never transmitted.
04. Security Architecture
We employ a Zero-Knowledge approach to authentication. All passwords are cryptographically hashed. We do not store "names" or any plaintext credentials in our databases.
05. Data Control & Deletion
Users have the right to purge their data at any time. AI-generated results can be deleted manually within the app settings. For a complete account-level wipe, users must contact their Digital Safety administrator. Once deleted, data is permanently removed from MongoDB and Vercel Blob.