A long-tail intrusion is a network intrusion that occurred in the past and has persisted for some time.
Long-tail intrusions include advanced persistent threats (APT). However, sometimes intrusions achieved by simpler methods and malware may be used as the starting point of a long-tail intrusion and continued by downloading new code by means such as Trojans hidden by root kits and code targeting zero day vulnerabilities.
Hackers of all stripes use these methods to gain long-term access for web crime, legitimate law enforcement, government spying and hacktivism. Access to the actual targeted data or network resource may take some time, during which the attacker could get other information from and about the network that may be valuable in itself both before and afterwards.
Long-tail intrusions are a considerable problem in security, especially for those relying solely on tier 1 security such as a firewall or intrusion prevention, which is focused solely on preventing initial entry to a network. In organizations with 5000 computers or more, 90 percent have existing breaches, on average, and these breaches may exist undetected for more than a year.
Tier 2 security such as breach detection systems (BDS) focus on malicious activity within the network it protects. BDS determines possible breaches by differing combinations of heuristics, traffic analysis, risk assessment, safe marked traffic, data policy understanding and violation reporting. BDS are sometimes able to detect breaches as they occur but the systems are designed to identify existing intrusions.