GUIDE EUROPE · DIGITAL SOVEREIGNTY

ENISA guidelines on adopting the Zero Trust model in European companies

Zero Trust drops the assumption that anything inside the corporate network can be trusted. Each access request should be checked against identity, device, context and the resource requested. This guide sets out the operating principles shared by European guidance and shows how to start with least privilege, strong authentication, segmentation and monitoring, introducing each control in phases.

Updated october 2026
7 min read
Altovar Technical Editorial Team

1. What Zero Trust is and what it is not

Zero Trust is a model that abandons the idea that everything inside the corporate network is trustworthy. The old perimeter, with a firewall at the entrance and full trust inside, no longer holds: users work from home, use personal devices and access cloud applications, so the real boundary runs through identities and devices. Every access request must be assessed on the basis of identity, device, context and the resource requested, and not simply because it comes from the internal network.

It does not mean buying a product or replacing everything at once. Zero Trust is a direction of travel that is introduced step by step, starting with the most sensitive data and applications. It is a good habit to be wary of the term as used in marketing: ask the vendor which concrete controls it offers, how it configures them and how it verifies them, not just what its solution is called. A promise without operational detail is worth little.

2. The operating principles shared by the guidelines

The guidance on the subject, including European publications, converges on a few principles: verify every access explicitly, apply least privilege, assume that a breach may already have occurred, and limit the damage through segmentation and monitoring. These ideas are simple, but the difficulty lies in applying them consistently to users, services and devices, especially when the tools have grown over time without a unified design.

For an SME, the most useful principle to start from is least privilege. Many companies have administrator users on every computer, accounts shared between colleagues and services with generic permissions. Reducing these privileges is an inexpensive measure that lowers risk immediately, without any need for new infrastructure. The work consists mainly of listing who needs what and removing permissions that nobody uses any more.

3. Identity and authentication: the new perimeter

Identity has become the real perimeter. Whoever authenticates the user, the device and the context is what really decides who gets in. That is why multi-factor authentication on email, remote access and cloud applications is the first measure to take: it substantially reduces the risk from password theft, which remains one of the most common routes in, especially when the same credentials are reused across several services.

Watch out for two mistakes. The first is accepting weak authentication for convenience, with exceptions that remain active for years. The second is forgetting service accounts and API keys, which often have no controls and are not covered by periodic access reviews. A technical account with broad permissions and a password that has never been changed is one of the most frequent starting points for an incident. The table summarises the areas to cover.

AreaObjectivePractical example
User identityKnow with certainty who is accessingPersonal accounts, no shared accounts
Strong authenticationMake credential theft more difficultMulti-factor authentication on email, VPN and cloud applications
Device statusAccess only from managed and updated devicesCompliance check before access to resources
Network segmentationLimit an attacker's lateral movementServers, printers and workstations in separate network zones

4. Segmentation: limiting the damage

Segmenting the network means dividing it into zones, so that a compromised computer cannot freely reach all the others. Microsegmentation takes the concept down to individual applications or workloads. An SME does not need to start at the finest level: separating at least servers, guest environments, workstations and production devices is enough. Even a simple separation, if done well, makes it much harder for malware to move from one machine to another.

A common mistake is creating segmentation rules without knowing the real traffic flows. Before blocking anything, observe traffic during a period of normal activity and understand which connections are necessary. Otherwise the rules break applications, employees ask for exceptions and the policy ends up being bypassed. Every exception must be documented with a reason and a review date, otherwise it becomes permanent.

5. Continuous monitoring and response

Zero Trust requires visibility. Without centralised logs you cannot verify who accessed what and when, or reconstruct an incident after it has happened. It is worth collecting authentication, network and device events in a single place, keeping them for a period defined by the obligations that apply to the company. Useful logs are the ones people actually read: too many unfiltered events generate noise and alerts that get ignored.

Monitoring is only worthwhile if someone acts on the alerts. Define who assesses an alarm, within what time and with which actions. For organisations subject to NIS2, incident management is a legal obligation; for others it is good practice that avoids discovering a breach weeks after it happened. Even a simple written plan, with the contacts to call, greatly reduces response times.

6. Getting started: a phased path

A realistic path starts with an inventory of identities, devices, applications and data. You do not need to complete it in one go: what matters is knowing where you stand and in what order to act. Priority goes to systems that hold the most sensitive data or that are directly exposed to the internet, because those are where an error does the greatest damage in the shortest time.

Once the first cycle is complete, ask your IT provider to describe in writing how it applies these controls to your systems. A concrete answer, with examples drawn from your own environment, is the best indicator of the maturity of the offer. A provider that can say which accounts have administrative privileges and when they were last reviewed is on the right track. If the answer is generic, that is a warning sign to investigate before signing.

  1. Inventory users, service accounts, devices and exposed applications.
  2. Enable multi-factor authentication on critical access, starting with email and remote access.
  3. Remove shared accounts and reduce administrative privileges.
  4. Observe network flows during a normal cycle of activity before introducing segmentation rules.
  5. Centralise logs and define who handles alerts.
  6. Review the approach at regular intervals and document every exception.

Frequently Asked Questions

Are the ENISA guidelines mandatory?

Generally not: publications by the European agency ENISA are technical recommendations, not legal rules. However, they become a useful reference for demonstrating that appropriate measures have been adopted, and some rules, such as NIS2, require risk to be managed appropriately. To see what applies to you, consult the most recent version and compare it with your own scope.

Does Zero Trust mean removing the VPN?

Not necessarily. A traditional VPN often grants broad access to the network once connected. In a Zero Trust model, access is granted application by application, after verification of identity and device. Many companies gradually replace the VPN with targeted access, keeping protection for the services that still require it.

How much does moving to Zero Trust cost?

The cost depends on the starting point, the number of users and applications and the tools you already have. Many first measures, such as multi-factor authentication, privilege reduction and centralised logs, may already be available in contracts you have: check before buying. It is sensible to start with an inventory and estimate costs by phase, so you can weigh the value of each step against the risk it reduces.

Is Zero Trust suitable for a small business?

Yes, if it is applied in phases and in proportion to risk. You do not need a complex architecture: start with strong identity, least privilege and separation of your most sensitive systems. Smaller companies often see quick benefits, because they have fewer legacy exceptions to manage and fewer systems to reorganise at once.

What should I ask a managed security provider?

Ask how it verifies user identity, how it manages administrative privileges, which logs it collects and for how long, who responds to alerts and within what timescales. Also ask how exceptions are documented. A clear answer, with examples from your environment, is worth more than a list of features.

About this content

General informational guide. Regulations and technical requirements change over time: always check the official sources and assess your specific case with a qualified technical or legal advisor.

GET IN TOUCH

Talk to our team.

We'll show how Altovar integrates into your infrastructure. No commitment, no sales pitch. Just a technical conversation.

BOOK A DEMO →