Biometric Trust
Deepfake & Synthetic Media Analysis
Deepfake and synthetic media analysis examines submitted photographs, selfie images, selfie videos and verification videos for indicators of synthetic media, face replacement, manipulation, replay and unusual compression or manipulation characteristics.
What is analysed
The architecture is designed to analyse submitted media for detectable anomalies: synthetic generation artifacts, face replacement or blending, manipulation of facial regions, replay characteristics and unusual compression or manipulation patterns that may indicate tampering.
Risk-based assessment
Results are represented as an assessment — Low Risk, Moderate Risk, High Risk or Inconclusive — rather than a binary verdict. Automated deepfake detection is a powerful risk indicator, but it is not infallible, and high-stakes decisions are escalated to expert review.
Distinct from capture integrity
Deepfake analysis examines the content of the media. It is deliberately distinct from capture integrity analysis, which examines how the media was captured — for example whether a virtual camera, replayed video or injected stream was used. Both assessments appear separately in the verification report.
Result States
Results are expressed as clear, technically defensible verification assessments:
Learn More
What is deepfake detection?
Deepfake detection analyses imagery and video for indicators of synthetic generation, face replacement or manipulation.
Detection systems look for artifacts left by generative models and editing tools — blending boundaries, unnatural texture patterns, inconsistent lighting and abnormal compression characteristics. Because generation techniques improve continuously, detection results are expressed as risk assessments within a broader evidence-based verification, never as a guarantee of detecting every synthetic media file.
Privacy note: this capability follows our privacy-first biometric architecture — explicit consent, data minimization, least-privilege access and audit logging. Biometric information is never exposed on public verification pages.
Related Biometric Capabilities
Get Started With Deepfake & Synthetic Media Analysis
Start a consent-based verification workflow and receive an evidence-based result.