Behavioral biometrics

Recognizing someone by how they type, swipe, and hold the phone.

Everyone has habits, such as how fast they type, how they move a mouse, and how they tilt a phone, and a system can learn those habits and notice when they change.

Continuous authentication from interaction telemetry such as keystroke dynamics and motion sensor patterns. Telemetry is the stream of measurements a device reports about how it is being used, and continuous authentication means the system keeps checking throughout a session instead of only at login.

A behavioral biometric system typically works like this:

1. The system collects interaction signals during sessions.
2. It builds a profile of how the genuine user normally behaves.
3. During a new session, it compares live behavior with the profile.
4. A large difference raises a risk signal, which can trigger a step-up check or a block.

Common signals include:

• Keystroke dynamics, such as typing speed and the pauses between keys.
• Mouse and touch patterns, such as cursor paths, swipe speed, and tap pressure.
• Motion sensor data, such as how the phone tilts and moves in a hand.
• Navigation habits, such as how someone moves through screens and forms.

Behavioral biometrics sit next to terms that are easy to mix up:

• Device fingerprinting recognizes the device, while behavioral biometrics recognizes how the person uses it.
• Bot attack is automated traffic, which behavioral signals can often tell apart from human use.
• Step-up authentication asks for an extra check, which a behavioral risk signal can trigger.