Our Belief
Privacy is the foundation of VeryAI's system
Pseudonymous by Design
Irreversible and Encrypted
Local-First Processing
No Data Retention
Verification
OAuth2
VeryAI implements OAuth2 for secure and interoperable authorization with partner applications. All access is consent-driven, requiring explicit user approval. Tokens are strictly scoped and time-limited, ensuring partners receive only the permissions necessary for their function.
Zero Knowledge Proofs
Zero-knowledge cryptography underpins VeryAI's trust framework by enabling identity verification without revealing biometric data. During palm verification, the user's app interacts with VeryAI's server to validate the scan and receives a digital credential. This credential generates zero-knowledge proofs that confirm identity while keeping biometrics hidden and unlinkable to a physical identity. Partner applications can verify these proofs independently, ensuring secure authentication without compromising privacy.
Revocation
Users can revoke their VeryAI account at any time. Revocation requires a palm scan to confirm ownership. Once verified, the account is permanently deleted and the associated biometric data is removed from the system. The user's Very ID is archived and marked as expired.
Expiry
Inactive accounts are subject to automatic expiration after three years. Users are notified and prompted to verify their identity through a palm scan. If no response is received, the account is canceled, and all biometric data and the Very ID are permanently removed to comply with GDPR.
Security
Biometric systems are vulnerable to spoofing attacks, where fake samples such as photos or video replays are used to gain unauthorized access. To counter this, VeryAI has built strong anti-spoofing capabilities at two layers.
Scarcity of Palm Data
Unlike facial images, which are widely available online, high-quality palm images are much harder to obtain. This scarcity makes the palm an inherently more secure biometric modality.
Gesture-Based Liveness Detection
VeryAI has developed a proprietary system that uses random hand gestures to detect and block fake biometric scans, including replay attacks.
