Behavioral biometrics are risk signals derived from how a person uses a device, such as typing rhythm, swipe speed, mouse movement, touch pressure, or navigation patterns. Unlike static biometrics like a fingerprint or face scan, these signals are continuous and probabilistic: they help estimate whether the current session looks like the legitimate user.
This matters because many attacks happen after login. Account takeovers, malware-driven fraud, and remote support scams can all reuse a valid password or session token while the attacker behaves differently from the real user. Behavioral systems compare live activity with a learned baseline and can raise risk scores when patterns change sharply. Defenders use them for step-up authentication, session monitoring, and fraud detection. They are not a standalone identity proofing method, but a layered control that helps catch abuse when credentials, device trust, or even other biometric checks have already been bypassed.