Incode Launches On-Device Age Estimation to Keep Facial Data on Users’ Devices

· Original article ↗

Summary

Incode says its age-estimation and passive-liveness models run locally, sending only an age-threshold result; server-side session metadata supports fraud detection. The sponsored article also describes Identiq fraud collaboration without pooled customer data.

Key points

  • Incode says its facial age-estimation and passive-liveness models run on users’ phones, tablets, or laptops, without transmitting or storing facial data.
  • The platform receives the age-threshold result; if the check cannot be completed, the user is offered another verification method.
  • Incode says it compressed the models to roughly one-tenth their original size using knowledge distillation so they can run in browsers or apps without special hardware.
  • The company says server-side session metadata—not facial or biometric data—is analyzed to detect camera-feed injection and device tampering.
  • Incode reports 99% spoof detection across deepfakes, injection, replay, and physical spoofing, and says its platform flagged more than 1 million face attacks in 2026; these are company claims.
  • The article says Identiq’s privacy-enhancing technology lets organizations share fraud signals without pooling customer data in a central data lake.

Article Details

Event Type
Incode announced a $100 million commitment to privacy-preserving identity infrastructure, acquired Identiq, and launched On-Device Age Estimation.
Impact
Incode says facial age estimation and passive liveness detection run on the user’s device, so facial data is not transmitted or stored. Its server-side layer uses session metadata for fraud detection, while Identiq’s technology is intended to enable fraud-signal sharing without pooling customer data.

People

Vendors

Products

Countries

Industries