BixeLab test shows Oz Forensics’ biometric IAD protects against deepfake fraud

Oz Forensics has confirmed the effectiveness of its injection attack detection (IAD) technology with testing by BixeLab.

BixeLab used the EU’s CEN/TS 18099 technical specification, which is also the starter document for an ISO standard in development.

The evaluation took place from May to August of this year, using Oz Forensics’ Web SDK, Android SDK and iOS SDK. The results were the same for all three: a zero percent APCER, meaning all attempted injection attacks were blocked, and 0 percent BPCER, according to a LinkedIn post from the company.

Injection attack method (IAM) checks used during the evaluation looked for virtual cameras, emulators or rooted devices and observed the network and API. BixeLab deployed injection attack instruments (IAIs) like static images, passport photos, prerecorded and deepfake videos, live video streams and face-morphing outputs to defeat the software.

IAD is widely used as a proxy for deepfake detection in biometric identity proofing and verification, identifying the method of delivering the fakes.

Face deepfake detection is forecast to generate $2.52 billion in revenue by 2027, according to the 2025 Deepfake Detection Market Report & Buyer’s Guide from Biometric Update and Goode Intelligence. The report includes a case study featuring Oz Forensics and a profile of the company.

Oz Forensics recently refreshed its compliance confirmation for both Level 1 and 2 presentation attack detection (PAD) with evaluations by BixeLab.

BixeLab is also the lab that partnered with MOSIP on the development of its biometric device certification framework, which includes an injection attack detection component.

Article Topics

biometric testing  |  BixeLab  |  CEN/TS 18099  |  certification  |  deepfake detection  |  deepfakes  |  injection attacks  |  Oz Forensics