Behavioral biometrics
Also known as: Behavioral biometric analysis · Keystroke dynamics · Typing dynamics · Passive biometrics
Recognizing a person by how they interact with a device — typing rhythm, cursor movement, how they hold and tap a phone — rather than by a physical trait like their face.
Legal basis
Wikipedia — Keystroke dynamics
Physical biometrics identify by what a person is (face, fingerprint). Behavioral biometrics identify by what a person does: the cadence of their typing, the pressure and angle of a touch, how they move the mouse, how they scroll. These patterns are hard to imitate and are captured passively, without asking anything of the user during the session.
In fraud prevention it serves two ends. It continuously confirms that it is still the account owner throughout a session — useful against account takeover — and it flags signs of coercion or automation, such as the hesitation of someone being coached by a scammer on the phone, or the mechanical regularity of a bot. It is a risk signal distinct from the device and the document, and it adds to them.
Frequently asked questions
How is behavioral biometrics different from facial biometrics?
Facial biometrics identify by a physical trait (the face). Behavioral biometrics identify by the way a person interacts with a device — typing, touch, cursor movement — captured passively during use, to continuously confirm it is the same owner and to detect coercion or automation.