AU10TIX is integrating Reality Defender’s deepfake detection into its identity-verification platform, adding a new signal for determining whether a selfie has been generated or manipulated with artificial intelligence.
The initial focus is remote customer onboarding, where an attacker may inject a synthetic image, use a face swap or otherwise manipulate the media presented to a liveness system. Reality Defender will assess the selfie for authenticity, and AU10TIX will combine that result with its existing document, biometric and fraud indicators.
AU10TIX will evaluate the authenticity result alongside document, biometric and fraud signals in its decision models. A single media classification will not serve as the only basis for an identity decision.
The integration addresses several forms of onboarding attack. Presentation attack detection addresses artifacts shown to a camera, while injection defense is intended to detect manipulated content inserted into the capture pipeline. Deepfake analysis adds another assessment of whether the face media itself is synthetic or altered.
Reality Defender has added its detection technology to other fraud platforms. Earlier this year, the company partnered with Charm Security to screen voice, image and text, extending its technology beyond a separate detection console into operational fraud workflows.
AU10TIX has also added continuous risk monitoring across the customer lifecycle and tools for authenticating non-identity documents used in know-your-customer and know-your-business checks.
The partnership comes after AU10TIX reported that AI-generated fraud had overtaken physical forgery in its network data. The company has argued that industrialized fraud operations increasingly reuse automated tools and synthetic media across multiple institutions. The partnership connects that network-level fraud data with media authenticity analysis during onboarding.
The combined workflow will compare media authenticity with document checks, liveness results and other fraud signals. Organizations can apply the resulting evidence through their existing onboarding thresholds and fallback procedures. The additional signal will sit alongside document authenticity, face matching and device intelligence during remote account opening.
The integration places deepfake detection inside the identity decision process. AU10TIX customers will be able to correlate media authenticity with document validity, liveness and wider fraud intelligence before approving an applicant.
Sources: Financial IT, AU10TIX
–
By the ID Tech Editorial Team
