Behavior-based security is a proactive approach to managing security incidents that involves monitoring end user devices, networks and servers in order to flag or block suspicious activity.
Traditionally, security management has been signature-oriented. In this approach, the security program monitors a data stream and compares source code in files or packets to the source code in an anti-virus vendor's library of known threats. In contrast, behavior-based programs compare the actions of files or network packets to a list of accepted or suspicious actions. In general, signature-based tools are best at identifying and repelling known threats, while behavior-based are best for fighting zero-day threats that have not yet made it onto a list of known threat signatures.
Behavior-based security software scans for deviations from the norm and has the intelligence to decide whether an anomaly poses a threat or can be ignored. Most behavior-based security programs come with a standard set of policies for which behaviors should be allowed and which should be considered suspicious, but also allow administrators to customize policies and create new policies.
A behavior-based security software product may be marketed as a behavior-based intrusion detection product, a behavior threat analysis (BTA) product or a user behavior analytics (UBA) products. Some products are sophisticated enough to apply machine learning algorithms to data streams so that security analysts don't need to program in rules about what comprises normal behavior. Others include behavioral biometrics features that are capable of mapping specific behavior, such as typing patterns, to specific user behavior. Most products have advanced correlation engines to minimize the number of alerts and false positives.