Using Incident Correlation to Reduce Cyber Threat Dwell Time

Attackers spend a considerable amount of time conducting reconnaissance on compromised networks to gain the information that they need to complete their objectives for criminal activity, including fraud and intellectual property theft. Dwell time, the amount of time an attacker is present in an enterprise is currently measured in the hundreds of days.

One of the most effective technologies available to incident response teams to help to reduce the threat actor dwell time and limit the loss of confidential data and damage, are Security Automation and Orchestration platforms. Security Automation and Orchestration technologies process alerts and correlates these with threat actors’ Tactics, Techniques, and Procedures. The ability to determine not only the initial ingress point of the attacker but any lateral movement inside the enterprise significantly reduces the time to deploy containment actions. In this scenario, the incident correlation engine is utilized not only as a mechanism for responding and orchestrating the response but also to proactively search for related IoC’s and artefacts. The synergy of response, automation and correlation provide organizations with a holistic approach to reducing cyber incident dwell time. In more mature organizations, these measures are leveraged frequently by IR responders to transition from being threat gatherers to threat hunters.

incman dwell time
Figure 1DFLabs IncMan Observables Hunter and Correlation Engine

When Incident correlation is available within the SAO platform, cyber threat dwell time is reduced through 3 separate but complementary capabilities:

  1. Category based correlation – Correlating incidents by type.
  2.  Asset based correlation – Contextualizing the criticality and function of an asset
  3. Temporal correlation -Providing insight into suspicious activity or anomalous access

Defense in Depth strategies is designed so that high-value targets, such as privileged accounts, are monitored for increased or suspicious activity (Marcu et al. 5). The incident correlation engine not only visualizes this but also provides information to help determine the source of an incident by identifying the points of entry into the affected infrastructure.

“Patient Zero” identification is accomplished through tracking the movement from a source to an end user, and assists responders in determining the epidemiology of the attack, and also possible intruder motives. The correlation engine can achieve this objective through correlating similar TTP amongst incidents and visualizing associational link analysis between hosts. This comparison produces a topology of the lateral movement and can easily identify and visualize the path of an intrusion and the nature of an attack. This permits incident responders to initiate containment actions in real time, as the intentions and objectives of hackers are readily determined.

Dwell time of cyber threats can be significantly reduced from the industry average length, currently measured in the 100s of days, to only a few hours by providing a system capable of identifying not only the magnitude of the attack but by providing a roadmap to successfully hunt the incident genesis point to prevent further proliferation.