Yoti confirms compliance for MyFace liveness tool with iBeta PAD retests

UK digital identity player Yoti has announced that its latest liveness detection tool, MyFace, has been re-tested by iBeta Quality Assurance and found to be in compliance with the ISO/IEC 30107-3 standard on biometric presentation attack detection (PAD, the legacy term for liveness).

According to iBeta’s confirmation letter, Yoti’s liveness technology was already confirmed at PAD Level 1 and 2 but Yoti requested the new test to ensure continued conformance. (Level 1 tests focus on commonplace objects, while Level 2 tests more advanced and specialized attacks; iBeta recently introduced a Level 3.)

In the lab’s words, Yoti’s latest evaluation was “a retest effort, involving a smaller number of artefacts and requiring less time than a full Level 1 or Level 2 PAD test process.” Testing accessed MyFace through the native Safari browser on an Apple iPhone 16 Pro running iOS 18.0, supported by the backend server components. Test subjects were cooperative.

The verdict found sustained compliance in MyFace, and a blog from Yoti says the tool achieved 100 percent detection of attacks with 0 false positives, and 100 percent approval for genuine presentations (0 false negatives). Yoti credits the success in part to an easy user experience; its liveness detection process is done with a single selfie, and does not require any action, such as repetitive prompts or facial expressions, from the user.

German regulator lowers age buffer for Yoti FAE by two years

Yoti has also gotten a nod from Germany’s media regulator for youth, KJM, which has reduced the buffer age for Yoti’s facial age estimation product from 5 to 3 years – meaning that, to access content restricted to those over 18, instead of needing to be estimated as over 23 (18+5), users can now be estimated as over 21 (18+3).

A post on the change notes that “Germany has one of the most stringent regulatory environments in the world, in particular when it comes to accessing adult content online.” But internal and external testing have shown Yoti’s FAE to be continually improving in accuracy and performance. “Over time, independent evaluations by NIST, the ACCS, as well as Yoti’s July 2025 white paper have shown significantly lower error rates and smaller Mean Absolute Errors (MAE), particularly around the ages of 17-20,” the firm says.

Its current MAE for ages 13-17 is 1.1. As well, it touts a false positive rate (FPR) of 6 in 1000 for 13-17-year olds being estimated as over 21, and a true positive rate (TPR) of 99.4 percent for the same demographic being correctly estimated as under 21.

Hence KJM’s decision to reduce the required buffer.

Article Topics

biometric age estimation  |  biometric liveness detection  |  face biometrics  |  facial age estimation (FAE)  |  iBeta  |  presentation attack detection  |  Yoti  |  Yoti MyFace