CryptoLab, a cybersecurity firm specializing in homomorphic encryption, will unveil its new Encrypted Facial Recognition (EFR) solution at the RSA Conference 2025 in San Francisco. The system is designed to enable facial recognition to be performed while sensitive data remains encrypted throughout the entire process.
EFR leverages Fully Homomorphic Encryption (FHE), specifically the CKKS algorithm, which allows mathematical operations on encrypted data without requiring decryption. The approach ensures that facial templates and related biometric information remain secure, even in the event of data leakage. The CKKS algorithm, designed for computations on real-number data, is particularly well suited for vision AI applications. Its structure also provides resistance to future quantum computing threats, aligning with cryptographic standards under consideration by the U.S. National Institute of Standards and Technology (NIST).
According to CryptoLab, the system is capable of real-time processing of tens of millions of facial templates with millisecond-level response times, made possible through GPU acceleration. This scalability makes EFR suitable for deployment in large-scale environments such as airports, smart city infrastructure, stadiums, and enterprise-level access control systems. The technology offers accuracy comparable to conventional plaintext-based systems while eliminating the risk of biometric data exposure.
The company describes EFR as a solution for organizations seeking privacy-preserving facial recognition in high-risk or high-traffic settings. Potential applications include secure identity verification at payment terminals, facility entry systems, and smart government services. EFR is designed to be fully compatible with commercial deployment environments and will be offered under license to enterprises and developers following its introduction at RSA.
The launch comes amid rising global concern over the security and ethics of facial recognition technologies. Privacy advocates have criticized systems that store unencrypted facial templates, citing risks of identity theft and unauthorized tracking. EFR’s encrypted-by-design approach is positioned as a direct response to these concerns, allowing for secure biometric authentication without exposing underlying data to system administrators, service providers, or potential attackers.
Growing interest in facial recognition for secure access and automation has led to high-profile deployments in professional sports venues and public transit systems. However, CryptoLab’s target market focuses on sectors where strict privacy compliance and cryptographic security are paramount. The technology is designed to support regulatory compliance with privacy frameworks such as the European Union’s General Data Protection Regulation (GDPR) and evolving U.S. state laws, although specific certifications have not yet been announced.
CryptoLab’s leadership brings substantial academic and technical credentials to the field. CEO Professor Jung Hee Cheon is credited with inventing the CKKS algorithm, now widely recognized in the field of encrypted computation. Chief Scientist Damien Stehlé is a co-author of multiple post-quantum cryptography standards endorsed by NIST. CryptoLab has also received international recognition, including being named a Gartner Cool Vendor and securing over 21 billion KRW in venture investment. Its algorithms have been finalists in Korea’s national post-quantum cryptography competition, underscoring the company’s role as a global leader in secure computation.
—
April 24, 2025 – by Ji-seo Kim
