Norway’s financial sector is evaluating privacy-preserving biometrics using homomorphic encryption for template protection, with facial biometrics specialist Mobai leading technical work as part of a national push to counter rising fraud. Reporting indicates banks lost roughly 930 million NOK in 2023, and that AI-enabled spoofs and deepfakes are intensifying the problem.
The concept is to run biometric matching on encrypted data, combining fully homomorphic encryption with multi-party computation so that sensitive templates are never decrypted during processing. The model aims to keep cloud providers blind to raw biometrics while allowing banks to control decryption of the final decision. Norwegian research partners and European academics have explored similar FHE+MPC patterns for wallet and remote-verification use cases.
CryptoLab’s “Encrypted Facial Recognition” unveiled at RSA 2025 and earlier work from NEC on homomorphic matching performance. CryptoLab’s encrypted FR and NEC’s homomorphic FR, as well as our digest note on CKKS-based approaches.
If successful, Norway’s work could offer a compliance-friendly path that aligns with GDPR and anticipated EU rules by minimizing biometric data exposure during computation.
—
By the ID Tech Editorial Team
