Centered around the principle that nothing both inside or outside a network perimeter can be trusted without verification, the Zero Trust model is attracting more and more attention from organizations that struggle to prevent data breaches with traditional approaches. However, enterprises that want to embrace this model must be prepared for certain jettison practices based on embedded notions of a secure corporate network and trusted insiders.
What is the Zero Trust Model?
The term was coined in 2010 by Forrester Research to describe a model of security where any device or person attempting to connect to a network is basically treated as untrustworthy. Instead of using network location as the basis for granting/denying access to network assets, the zero trust model emphasizes the use of user and device credentials. Security experts say that this model is vital for preventing cyber-attackers from moving around inside a network undetected while looking for valuable targets after they’ve breached the perimeter. It’s when data leak prevention and security control tools are unable to detect malicious activities that breaches typically happen.
Where Is the Problem?
The problem is in the process, long-used by enterprises, of trusting traffic and users on the internal network, while the only untrusted users are the external ones. Next, the increasing use of cloud services and the growing mobile workforce are also making it harder for organizations to establish and enforce a secure network perimeter. Enterprise data has become increasingly scattered, there are many ways for accessing it, so the (traditional) perimeter-centric security strategies are not useful anymore.
The traditional network model divided users and clients into two groups – trusted and untrusted. The trusted sources were those inside the network, while the untrusted were external to it (they could be either partner networks or mobile users.) To access the internal system and recast to become trusted, one would typically use a VPN (Virtual Private Network).
Everything is untrusted in zero trust – there can no more be trusted people, trusted devices or trusted networks. Some enterprises segment network using VLANs (Virtual Local Area Networks) but what VLANs do is they isolate network traffic, but can neither enforce the control of sensitive assets nor inspect traffic for threats. Zero trust segmentation requires security that understands your users, content, and applications.
There is inadequate control, protection, and visibility of application and user traffic transitioning high-risk network boundaries. The assumption that everything and everyone on the inside of one’s network should be trusted is outdated.
Building a Zero Trust Network Infrastructure
With a zero trust network infrastructure, all users and data traffic will be assumed to be operating from an unsecured network. Thus, all network traffic will be encrypted. All of that makes the cybersecurity architecture a lot tougher because users will have to validate their credentials (with multi-factor authentication) every time they want to access the network. For enterprises still using the traditional perimeter defense, this may seem too difficult. However, this model is becoming more and more critical.
What does it take to build a zero trust network?
Zero trust network infrastructure must be designed based on its requirements. An enterprise will need to identify where the microperimetry are placed (depending on their data flow). A zero trust network infrastructure requires a structural change at the very beginning, which makes it quite difficult to incorporate it in the middle of an existing model.
Understanding the flow of data across their system is an excellent preparation before building a zero trust network. That allows organizations to understand which stakeholders require what kind of data.
Identification of sensitive data assets is the most critical step in employing the zero trust model. The crucial information that’s required is how the data stores, how sensitive it is, how it is used and by whom. After identifying these features, the data should be classified.
Access control is the most prominent component of the zero trust model. Enterprises must formulate policies about which users get what kind of access because access is provided on a limited and restricted need-to-know basis.
Zero trust networks require constant review and supervision of the network situation. In this type of network model, all traffic is critical (not just the external one.) Continuous monitoring is essential for finding the source behind all traffic with the purpose to flag any unusual activity inside the internal network.
By adopting the “always verify, never trust” motto, you will make sure that everyone accessing your enterprise’s sensitive data are authorized to do so. When it comes to cyber-attacks, the end user is often the point of access (whether by accident or intentionally.) But by getting ahead of cyber-attackers, you can prevent them from finding an entry point, so they won’t even get the chance to try to breach it. Thus, our networks need to be intelligent enough to allow only authorized and authenticated sources, because nothing should be trusted in today’s digital world.