Mitek Systems has announced significant enhancements to its SEPBLAC-compliant digital onboarding capabilities for the Spanish market, adding advanced defenses against injection attacks where fraudsters insert pre-recorded deepfakes directly into biometric verification data streams.
The enhanced solution builds on Mitek’s Digital Fraud Defender platform, launched in December 2024 to detect digital injection, template, and deepfake attacks. The update addresses the rapidly escalating wave of AI-driven identity fraud affecting Spanish regulated entities, with 89 percent of Spanish companies reporting increased fraud attempts annually.
Mitek’s updated unassisted video verification platform now incorporates four key defense layers: deepfake detection for AI-generated facial manipulation and face morphs, digital manipulation and screen-replay analysis to detect recorded or altered video content, injection attack protection with safeguards against malicious data insertion attempts, and face gallery analysis to identify repeated or coordinated fraud attempts.
The timing coincides with the February 2025 publication of a new European injection attack standard (CEN/TS 18099:2025) that provides frameworks for testing these defenses. Security researchers have documented iOS video injection tools that can bypass biometric checks by feeding deepfake streams directly into verification apps.
The company reports injection attack detection accuracy rates of 99.9 percent on known deepfake generation engines, with deepfake injection detection rates—including face swaps, diffusion techniques, and image animations—exceeding 99 percent.
“The threat landscape has fundamentally changed, and verification systems must evolve at the same speed,” Marc Sabadí, Innovation Lead and Sales Manager at Mitek Systems, said.
The enhancements are specifically designed for compliance with SEPBLAC (Spain’s anti-money laundering authority) requirements for unassisted video verification, which financial institutions and regulated entities must use when conducting remote identity checks. Mitek’s multi-layered approach combines an AI layer trained to detect a wide range of injection artifacts with traditional software-based checks, monitoring both digital content and data streams to identify visible and invisible signs of injected fraudulent content.
Identity fraud is now the leading cause of corporate fraud losses in Spain, with most attacks targeting the onboarding process itself rather than post-enrollment stages. The solution is designed to balance fraud prevention with seamless user experience for legitimate customers.
–
By the ID Tech Editorial Team
