Artificial intelligence is advancing at breakneck speed, and biometric authentication with liveness detection—once considered the gold standard in digital identity verification—is now under siege. Deepfakes, synthetic media, and AI-generated spoofing tools are more accessible and convincing than ever. Traditional systems relying on cloud-based analysis or static liveness checks are dangerously outdated.

Deepfakes, synthetic media, and AI-generated spoofing tools are more accessible and convincing than ever. Traditional facial recognition systems, especially those relying solely on cloud-based analysis or passive liveness checks, are completely obsolete, despite their prevalence in fintech, DeFi, and digital banking worldwide. At the same time, threat actors no longer need sophisticated tools to bypass standard facial recognition systems. A free, anonymous email account, some AI video gen software off the internet, and a still image or two from any social media account are now enough to fool most identity verification platforms – this is because they do not process the data locally.

The Threat

• AI-powered fraud now makes it possible to bypass many KYC onboarding processes with nothing more than a still image, a free email account, and widely available deepfake software.

• Cloud-based verification platforms introduce additional risk—sending sensitive biometric data offshore, often to vendors with questionable ownership, opaque data handling, or ties to jurisdictions that undermine privacy and sovereignty.

• Fintechs and DeFi companies face heightened exposure, especially when relying on providers in the UK, US, Canada, and EU that use offshore subprocessors or outdated verification models.

Most systems labeled as “liveness detection” perform only surface-level checks before sending the image to the cloud for advanced processing. This forces them to rely on outdated 2D image processing often provided by questionable offshore data processors, making them easy targets for presentation attacks using photos, deepfake videos, or even AI-generated avatars. Biometric systems that were once built to stop fraud are now frequently bypassed by it.

“AI-driven fraud is exploding across legal, real estate, and financial services. This is a technology arms race. The only way to win is to meet AI with better AI, backed by privacy-first architecture. With our edge-computing biometrics, your users’ most sensitive data never leaves their device, and fraud attempts never reach your systems.” said Matthew Unger, CEO at iComply

The iComply Platform: Built for the Next Era of Threats

We’ve spent the last five years engineering and refining a better Live Face Match biometric authentication system that can perform any type of check directly on the user’s device. This not only addresses these modern threats, it is a game changer for personal data privacy and national data sovereignty. Our latest release of the iComply platform delivers randomized, concurrent liveness and biometric testing. Performed entirely on-device via our proprietary edge computing architecture to detect and neutralize generative AI spoofing before it can infiltrate your onboarding process.

Fall 2025 Release Highlights

1. Advanced Multi-Expression Live Face Match Testing: Enhancements to performance and concurrent processing of both biometric face matches and liveness detection algorithms. Our platform doesn’t just check for motion and a face match; it challenges users to perform randomized facial expressions and micro-movements in 3D, making it nearly impossible for pre-recorded or deepfaked media to replicate. Each expression is evaluated independently alongside biometric confidence scores and device metadata to create your confidence threshold, which can be customized based on your risk tolerance.

• Real-time 3D facial recognition combined with randomized micro-expression prompts.

• Concurrent biometric and liveness analysis makes pre-recorded or AI-generated forgeries virtually impossible to pass.

• Independent scoring for each challenge, combined with device metadata, allows for fully configurable pass/fail thresholds.

2. Edge Computing for Real-Time AI Fraud Detection: Unlike API driven KYC or identity verification systems, our identity and biometric checks are performed directly on the user’s device through edge computing. Edge-computing ensures your customer data is always processed locally, in the country where they are at that moment, and validated before you touch it. This reduces exposure, accelerates processing time, and ensures biometric data never leaves the device, drastically improving both privacy and security. With this release, Pro and Enterprise accounts can now leverage enhanced configurability and data localization control for emerging regulations covering data privacy, security, and sovereignty.

• All biometric processing happens locally, on the user’s device. This ensures that data never leaves the country of origin. Zero data leakage. Zero third-party processing.

• No reliance on offshore cloud processors means significantly reduced attack surface, zero transmission risk, and compliance with emerging data sovereignty laws.

• Enhanced configurability for Pro and Enterprise clients to meet national and sector-specific privacy mandates.

3. Enhanced Threshold Controls for Precision Matching: Manage thresholds for biometric confidence score, adjust pass criteria, and the number of facial expressions required to be completed successfully.

• Dynamically set biometric confidence thresholds (e.g., 70%, 85%, 95%) based on your risk profile.

• Adjust requirements based on the risk and use case of the biometric verification event.

AI Isn’t Going Away, But Neither Are We
Organizations can no longer rely on “good enough” systems from five years ago to stop the threats of today. AI-generated fraud is evolving faster than most compliance teams can adapt. Without advanced, on-device defences, organizations risk onboarding bad actors, breaching data protection laws, and undermining user trust. By engaging iComply as their AML compliance technology partner, our clients reduce cost, manual operations, and fragmented systems while gaining clarity, consistency, and confidence in their AML compliance program. A program that is built not just for today’s threats but also for the upcoming threats posed by generative AI and offshore data processing.

About iComply
iComply is a global leader in modular compliance solutions for KYB, KYC, KYT, and AML. Founded in 2017 and headquartered in Vancouver, Canada, iComply helps regulated and emerging financial services providers operate with trust, accountability, security, and privacy. Our proprietary edge computing technology processes and encrypts sensitive identity data directly on the user’s device, enabling compliance without compromising privacy or data sovereignty. The iComply platform consolidates up to eight legacy vendors into one secure, configurable system—reducing compliance costs by up to 90%, improving customer satisfaction by over 25%, and ensuring readiness for evolving regulations in over 195 countries and 142 languages. Learn more at www.icomplyis.com.