Startup Claims 99.999% Accuracy for ‘Implicit Authentication’ System

A new startup is offering a novel approach to user authentication based on passive behaviors that are tracked intermittently via mobile and PC devices.

The company is called UnifyID, and is emerging out of stealth mode from the Stanford StartX incubator. Its eponymous platform is aimed at leveraging what the company calls implicit authentication—a collection of unique identifiers collected passively from normal user behavior.

Many of these are behavioral biometrics. On a laptop or desktop, the could include typing speed and mouse or touchpad movements; while a mobile device’s built-in accelerometer and gyroscope could recognize a user’s gate—and that in only four seconds, according to the company. This information is combined with metadata such as GPS location and WiFi signal, and UnifyID says that it can attain an identification accuracy of 99.999 percent using only four metrics, though the platform has more than a hundred such factors at its disposal.

It’s a concept similar to one being explored for Google’s Android, with that IT giant aiming to establish a passive mobile authentication system that could eliminate the need for password-based security by the end of this year. And it’s an approach that brings technological authentication closer to the kind of recognition we use with the people we know every day—not asking for a password or fingerprint, but just seeing and hearing our acquaintances for who they are.