In the era of a highly mobile workforce, endpoints offer access into corporate resources from virtually anywhere in the world. While this is great for offering flexibility to hybrid and remote workers, this also increases the opportunity for an attacker to compromise an endpoint and gain access. In this article, we'll cover why foundational endpoint security is important, what the challenges surrounding it are and how Idira Endpoint Privilege Manager (EPM) can help you go from “zero” to immediate risk reduction on day one to better prepare you for implementing your endpoint privilege management controls quickly at scale.Introduction
A common cyber defense system will include threat mitigation tools such as Network Firewalls, Intrusion Prevention Systems (IPS) and Endpoint/eXtended Detection & Response (EDR/XDR). These tools, however, are all reliant on detection of “known bad” signatures or behaviors. While this type of protection is a key part to a holistic cyber security program, it can be hit or miss against targeted or zero-day threats. To address this limitation, events are collected and correlated to try and discover indications of an attack using Security Information and Event Management (SIEM). This is a time-consuming process which is prone to false positives and negatives, all the while giving malicious threat actors more dwell time to execute their attack.How much damage is done during that time is determined by how good an organization’s foundational security is. This could make the difference between your detection technologies being an effective way to hunt down attackers while looking for weaknesses, or them providing you a front row seat to the horror show.What is foundational endpoint security?
These are the controls you have in place to harden endpoints and negate or minimize the damage that an internal or external threat actor can inflict. Foundational endpoint security is not a new concept, and the advice from leading authorities has been very consistent over the years (e.g. SOX, PCI, HIPAA, etc.). Time and time again we see Application and Administrator Rights Control at the top of the lists for foundational security. The “Pyramid of Cyber Security Needs” builds upon Microsoft's “Hierarchy of Cyber Needs” concept. Here we can compare the importance of the controls against the amount of resources typically allocated to each control. The effectiveness of each layer is dependent on the previous layer. For instance, how can we implement effective access controls if we can't trust the legitimacy of the identities we are granting them to.
Figure 1: Pyramid of Cyber Security Needs Diagram
The Challenge
Foundational layers such as “Access Control” and “Software” are often perceived as difficult to implement, leaving them under resourced for many organizations. Common reasons for this include, but are not limited to, barriers such as environmental complexity, fear of impacting business operations, and burdens related to high management overhead (whether that is real or perceived). This can lead to large gaps in security that will leave an organization vulnerable once an attacker gets past the external line of defense. If a zero-day threat makes it past an organizations detection systems and foundational security controls aren’t in place, the attackers will inherit permissions of whatever accounts they compromise. This leaves organizations vulnerable to user impersonation, ransomware attacks, or even worse, total network takeover.The Solution
Idira addresses these challenges by using proactive application and privilege management controls in an easy to deploy manner. By adhering to the recommendations below, you’ll be able to go from zero on the scale of foundational controls, where end users are local admins and machines are vulnerable to ransomware & and malware, to immediate risk reduction on day one. This is all helping us work towards our ultimate goal of achieving a full least privilege program where only known and approved applications are able to run and perform privileged actions.To remediate the risk associated with access control and software, we’ll want to apply a combination of least privilege controls on the endpoint that affect the user themselves and any other software or services running on that machine. For the controls themselves, we’ll implement least privilege role-based access for the user on that machine, removing admin rights and elevating as needed with full audit. This helps address privilege escalation and abuse risks by ensuring users can only perform the functions required on that system per their role or job. We’ll also be implementing smart application control to quickly mitigate the credential theft, lateral and vertical movement and privilege escalation and abuse risks, helping us work towards our goal of only allowing trusted software to run while preventing malicious software from executing.Least Privilege Role-Based Access
There may be some debate as to whether Application Control and Allowlisting should be prioritized before controlling Admin rights, but the truth is that the two are inextricably linked. The privilege management capabilities in operating systems are very simple. You are either an administrator, with unlimited access or you’re a standard user. Wherever users are operating with local admin rights, we are never truly in control of that endpoint. Any security tools you put in place, including application control, can be bypassed by an administrator or an attacker impersonating an administrator.Microsoft recognized this was a problem early on and created some additional static roles such as Power Users, Server Operator, and Backup Operator. However, these are largely non-functional since they provide too few rights to be operational, while simultaneously enabling overreach in other areas. This leads us back to the original problem. If a user needs admin rights for a particular application, script or function (however small), the only realistic option provided is to grant them full admin rights. This means that these privileged users and any attackers impersonating them can:
- Deploy system wide and persistent malware
- Install any desired software
- Disable and uninstall security software, services, or policies
- Tamper with system configuration and OS files
- Perform user account attacks on any account that has ever logged on to that endpoint
This enables them to move laterally and increase the scope of their attack; and this is just a short list of what can be done. By design, administrators have unlimited access. We can overcome these obstacles by implementing least privilege controls in the following way:
- Remove users from the Local Admins Group so that all users will operate from a standard user account.
- Start applying controls and enable instant auditability around these accounts from day one.
- Then we can define roles tailored to your business in order to empower users with the specific rights they need at the OS and application level.
This is all accomplished without fundamentally changing the user experience. The best part is - these roles don’t need to be static. We can also apply context here as well. Let’s take a closer look at the diagram below.
Figure 2: Least Privilege Role-Based Access Diagram
When a Developer Logs into their laptop while at home, their Resultant Set of Policy (RSOP) will be evaluated as follows:
- Developer rules always apply as that condition is constant, and based on their role
- Remote Worker rules now come into play based on conditions such as network access and available resources
- Baseline Policy rules (e.g. rules that apply to the larger organization) apply next along with any Break glass/Just-In-Time rules that have been approved.
Finally, anything not handled by policy will fall into discovery and exception handling, which you can align with existing approval processes such as helpdesk ticketing. This is a key component to discovering new use cases and keeping your role-based policies current with your users ever changing needs.Smart Application Control
The next key element to foundational security is Application Control. Most organizations really struggle with this as traditional app control requires you to allowlist every known good application before realizing any meaningful protection. Even then, attackers can find ways around it. On the other hand, smart app control addresses that challenge and enables us to achieve end state results very quickly. But what does smart app control actually consist of? It’s a combination of targeting known and common attack paths of malicious actors, including content handlers, credential stores and ransomware. Keep in mind that all of the following concepts are quick ways to get measurable risk reduction without affecting the end user experience. As we continue to discover more insights around end user behavior, we want to continually refine these policies to work towards a full-least privilege model.Securing Content Handlers
Built-in tools are often weaponized for fileless attacks, in an approach that is referred to as living off the land. This allows an attacker to bypass traditional app control as the apps being used will be allowlisted. Your users legitimately need them, so an attacker is also able to misuse them. By invoking context, we can prevent built-in tools as well as unsigned/unknown applications from being executed by apps that are commonly exploited - in this case content handlers.
These apps don’t need that functionality, but they inherit it from the operating system, and it’s frequently weaponized through the use of malicious content. This is a huge risk reduction that doesn’t require any knowledge of the payload and can be implemented without impacting the user experience.Credential Theft Protection
When the credential theft protection feature of EPM is enabled, only approved and relevant applications (which will be a short list) can access credential stores or caches (such as the LSASS process, Chrome password store, etc). Anything else will be denied, or you can optionally offer up credential lures instead as a honeypot for detection. Attempts to access or use credential lures will set off an alert chain, allowing security teams to zero in on internal and external attackers attempting to gain a foothold. This is where EPM and detection technologies like EDR can supplement each other and work better together.Ransomware Protection
Similar to credential theft protection, ransomware protection enables us to restrict content access only to legitimate applications such as office and other productivity suites. To configure this policy all we need to know is the content handlers that you commonly use, which again will be a short list of trusted apps. Any undefined apps (including scripts, child processes etc.) will be prevented from tampering with content. This is a simple, but extremely effective security measure that again requires no knowledge of the payload. SMART APP Control allows us to think differently about how we defend endpoints, allowing us to focus on what we want to protect or prevent and work backwards from that.Fearlessly Forward with the Foundations
While there are certainly many worthy security initiatives to tackle, we hope the insight and guidance above helps address the risks associated with endpoint security. Privilege management and application control on endpoints are a fundamental element of an organization’s IT security hygiene but are commonly overlooked.Stay tuned for our upcoming parts in this series as we take a deeper dive into implementing these foundational controls. In the meantime, we encourage you to check out our videos on Endpoint Privileges Manager.