Protecting User-Endpoint Interactions

Stephanie Gray

Blog

25 October 2018

Earlier this month the CrowdStrike® Falcon® Overwatch™ team released their 2018 mid-year review, “Observations from the Front-Lines of Threat Hunting”. InfoTrust discusses the front-line and why security is everyone’s business. A brief precis, some thought provocation, and insight (hopefully) are below.

What is the front-line?

In security there is an adage, the endpoint is the lowest common denominator within an enterprise environment.  Others take this a step further quoting “users are the weakest-link”.  Regardless of where your staff work, or how many endpoints an individual interacts with, the interaction between users and endpoints should be considered your front-line. Your users can change the way your endpoints are used and with the same principle your devices can alter the way your users behave.

Modelling user-endpoint interactions as a core security perimeter require businesses to migrate from traditional security controls and concern themselves with identity relations and investigative tools.  Managing the interaction between user-endpoints across the organisation can start with user awareness training, identity, and privileged access management (IAM and PAM respectively), and dedicated endpoint detection and response tools.  Understanding is required for greater control.

This user-endpoint interaction is a dynamic corporate perimeter that requires input from security, information technology, and human resources professionals.

Who’s on the front-line

Whilst IAM and PAM apply authorisation, authentication, and accounting to your user-endpoint interaction, understanding and examining the interaction requires an additional skill-set.  Numerous systems can monitor and alert for a given rule-set built upon the accounting principles of your IAM and PAM, however, critically assessing user-endpoint patterns, determining suspicious activity, and investigating beyond is the remit of a threat hunter.  To use an additional metaphor common within InfoTrust, threat hunters find the needle in the haystack by filtering through the hay.

This is where the CrowdStrike® Falcon® Overwatch™ team come to the fore, monitoring the user-endpoint relationship, wherever they are, and detecting “intrusions by sophisticated or persistent adversaries that might otherwise go unnoticed”.  The OverWatch™ team actively hunt through CrowdStrike deployments across the world, inspecting and exploring any potentially suspicious event, any behaviour that is out of the ordinary for a user-endpoint relation.

Monitoring the user-endpoint interaction requires tools that look at behaviours and allow for threat hunting by retaining useful data for later analysis.  An example of this is the CrowdStrike® Falcon® agent, securely reporting back to the Falcon® platform where OverWatch™ can perform investigations and make informed decisions.

Key Observations & ATT&CK Framework

The first half of 2018 saw the introduction of the European GDPR and the Australian Notifiable Data Breaches (NDB) scheme.  In the first quarter of the NDB, there were 242 notifications with 143 of these due to a malicious or criminal attack. The research also found that 77% of all incidents were due to some form of compromised credentials.  The timeframe of the mid-year OverWatch™ report overlaps with the OAIC NDB quarterly report, also showing that compromises involving both user and system credentials were most prolific.

CrowdStrike® Falcon® OverWatch™ mid-year review maps incidents to the Mitre ATT&CK framework, allowing analysis of advanced attacks.  The framework determines phases of an attack into eleven categories, which the OverWatch™ team has embedded into the CrowdStrike® product.

• Initial Access
• Execution
• Persistence
• Privilege Escalation
• Defence Evasion
• Credential Access
• Discovery
• Lateral Movement
• Collection
• Exfiltration
• Command & Control

The OverWatch™ report shows the most common attacks methodologies included initial access typically from spear phishing, either via link or attachment.  Then defence evasion and privilege escalation using valid accounts and, once the attacker was in the network, statistically, they would perform either account or user discovery using credential dumping.  This correlates with the OAIC NDB report which showed that attacks focussed on credentials.

Behind the adversarial tactics and techniques is a clear interaction between accounts and systems, users and endpoints, recommending an increase in user-endpoint security.

What can you do?

Phishing and compromised credentials were the core methodologies from the 2018 Falcon® OverWatch™ mid-year review, both central to user-endpoint interactions.  InfoTrust recommends organisations take a few basic measures to tighten their perimeter and protect the front-line with technology already in place.

• Implement impersonation controls on your secure email gateway (SEG) where possible (or evaluate technologies that utilise machine learning models to augment traditional SEG solutions)
• Establish information risk management policies and controls for deployment of corporate and administrator user accounts. Recommendations such as segmenting administrative functions from the rest of the network.
• Roll out a security awareness program across your organisation; incorporating security awareness videos and workshops

However, these basic controls are only one facet of a security in-depth, and in-breadth, practice.  Increasing your capacity beyond basic controls requires a paradigm shift, from traditional layered preventative controls to detective modelling and the gradual increase of threat hunting capabilities.  Solutions that model acceptable behaviour to detect that which is suspicious allow investigation and analysis across a broad range of use-cases layered over preventative approaches.

Tools that allow threat hunting by monitoring the user-endpoint ecosystem and applying behavioural modelling to email identities and endpoint communications provide greater insight into potential threats. Coupling this with advanced IAM and PAM solutions to segment your network more effectively only improves any behavioural modelling or machine learning implemented across your security architecture.

To find out more, talk with an InfoTrust representative today.

Endpoint Detection Response

Threat Hunting

Share