日本語 
ELEMENTS Inc. has cleared a major milestone in biometric security. Its face recognition spoofing detection AI, Liquid PAD, has received official confirmation for the international biometric authentication security standard ISO/IEC 30107.
The certification was issued after testing by Fime, a third-party evaluation organization. The results were as clean as it gets. Liquid PAD recorded a 0 percent false acceptance rate for spoofing attacks and a 0 percent false rejection rate for genuine users.
Liquid PAD sits at the core of ELEMENTS Group’s identity and authentication stack. The group operates LIQUID eKYC, Japan’s leading eKYC service, and LIQUID Auth, a cloud-based facial authentication service. Liquid PAD powers the face recognition fraud detection layer across both products.
For context, LIQUID eKYC is designed to automate identity verification using AI screening. The aim is simple. Reduce user drop-offs while keeping verification fast, low-cost, and secure. LIQUID Auth extends this into ongoing authentication, balancing usability and security. Liquid PAD is embedded in both, handling impersonation and presentation attack detection.
Also Read: SoftBank to buy DigitalBridge in $4 billion deal to boost AI infrastructure
Standard ISO/IEC 30107 was specifically established to deal with the testing and evaluation of biometric authentication systems. When a system is said to be cleared, it implies independent verification for the system against the practical risks of spoofing, and not merely laboratory-friendly scenarios.
Looking ahead, ELEMENTS is not treating this certification as the finish line.
Facial recognition threats are evolving fast. Beyond photos and video-based presentation attacks, the industry is now dealing with camera injection attacks and increasingly realistic deepfakes. Countering these requires layered defenses.
ELEMENTS is pushing R&D on multiple fronts. This includes strengthening device-level security to address injection attacks. On the deepfake side, the company is developing techniques to detect inconsistencies unique to AI-generated imagery. These efforts are being combined with face reuse detection and other mechanisms through Liquid Shield to tackle attacks that are difficult even for humans to spot.
Longer term, the goal is broader than biometrics alone. ELEMENTS plans to build more comprehensive fraud prevention solutions by combining biometric signals with multifaceted data and behavioral patterns.
The direction is clear. Fewer false rejections for legitimate users. Stronger resistance against sophisticated fraud. And a system where honest users are not penalized because of a small number of bad actors.
English