AI in Biometric Authentication

Ten years ago, video identification solved an important problem: how to confirm that a remote person is who they claim to be. In doing so, the technology helped unlock a world of remote services.

However, the rapid evolution of deepfakes and injection attacks has undermined both human ability to detect spoofs and the reliability of video identification as a secure method of identity verification.

Andrew Bud, Founder and CEO of iProov, explores:

• The evolution of injection attacks and deepfakes: Insights from iProov threat intelligence into the tools and techniques bad actors use to bypass remote identity systems like video identification.
• Human fallibility in detection: How deepfake quality has surpassed human perceptibility—only 0.1% of people can reliably spot deepfakes.
• How strong liveness can secure video identification: Liveness that confirms a remote user is the right person, a real person, and authenticating in real time is essential to secure against deepfakes.
• The need for adaptive defenses: Software alone will inevitably be broken. Liveness technology must be maintained by expert human oversight to adapt to evolving threats.
• Evaluating a vendor: How to select a liveness vendor using standardized tests, third-party certifications, and customer references.

As biometrics become central to digital identity, AI is redefining how we authenticate users — making systems faster, smarter, and more adaptive. But with innovation comes complexity: deepfakes, bias, and evolving attack vectors raise critical questions around trust and fairness.

This panel brings together experts to explore how AI is shaping the future of biometric authentication — from liveness detection and behavioral biometrics to privacy-preserving machine learning. Join us for a candid discussion on where AI is improving security, where it's falling short, and what it means for the next generation of identity systems.