Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
ZTNA connects a verified user and device to one named application rather than placing them on the network. It is what replaces the VPN in a zero trust architecture.

What Is ZTNA?
A VPN answers one question — are these credentials valid — and then places the user on the network. From that point they can reach whatever the routing and firewall rules allow, which in most organisations is a great deal more than their job requires. Zero trust network access inverts the model: the user authenticates to a broker, the broker evaluates identity, device state, and context, and then connects that session to one specific application. There is no network route, and nothing else is visible.
The user’s client, or their browser for web applications, contacts the ZTNA service rather than a network gateway. Authentication runs through the organisation’s identity provider, ideally with phishing-resistant multi-factor authentication. Device posture is assessed: is this a managed device, is disk encryption on, is the EDR agent healthy, is the operating system patched. Policy then decides what this user, on this device, in this context, may reach.
The architectural detail that matters most is the direction of the connection. A connector deployed beside the application establishes an outbound session to the broker and holds it open. Traffic is stitched together in the middle. The application therefore needs no inbound firewall rule and no public address, which means it cannot be scanned, enumerated, or exploited by anyone who has not already passed authentication. This is sometimes described as a dark network, and it is the single largest difference from a VPN in practice.
The VPN concentrator has become one of the most reliably exploited categories of enterprise equipment. The Ivanti Connect Secure chain of CVE-2023-46805 and CVE-2024-21887, exploited from December 2023 and the subject of an NCSC alert, allowed unauthenticated attackers to execute code on the device terminating trusted sessions. Fortinet, Citrix, and Pulse Secure have all produced equivalent incidents. Each one is an internet-facing appliance you must patch urgently and that grants broad access once passed.
The second driver is blast radius. Credential theft is the most common initial access vector; a stolen VPN credential yields a network foothold, while a stolen ZTNA credential yields one application, and continuous evaluation can revoke it mid-session when device posture changes.
The third is the one users notice: performance. A ZTNA client sends SaaS traffic directly rather than backhauling it through a data centre, and access is no longer all-or-nothing.
ZTNA secures access to applications. It does not secure the applications themselves — an authenticated user reaching a vulnerable web application still reaches a vulnerable web application, so authorisation inside the application remains essential.
It handles web and TCP applications well, and legacy protocols less well. Thick clients, applications that initiate connections back to the user, VoIP, and anything expecting multicast or broadcast will need testing and sometimes a residual VPN path. Plan for a period of running both.
Device posture checking on unmanaged devices is weak, because the agent reporting the state runs on the machine you do not control. Browser-based access with no agent is convenient for contractors and gives you correspondingly little assurance about the endpoint.
And the broker is now a concentrated dependency. Its availability is your availability, and it decrypts your traffic. Ask about failure behaviour, point-of-presence locations relative to your users, and where logs are stored, which under UK GDPR belongs in your records of processing.
Start with one application and one group — a contractor population, or an internal tool with a clear owner. Run ZTNA alongside the VPN rather than cutting over. Write policy in terms of identity, group, and device posture from the beginning, not in terms of addresses, because carrying the old model across is the commonest way these projects end up as an expensive VPN.
Expand application by application, measuring what each group actually needs rather than granting what they had before. Decommission the VPN only when the remaining use cases are genuinely covered, and put a date on it, because a VPN kept running indefinitely as a fallback stays as a parallel path with all its original risk.
ZTNA is the remote access component of a zero trust architecture as described in NIST SP 800-207, and it is usually bought as part of SSE. Internally it pairs with microsegmentation, which applies the same thinking to east-west traffic. Our Infrastructure and Architecture Review Services cover the migration path and the policy model it depends on.
A VPN authenticates once and places the user on the network, where routing and firewall rules determine reach. ZTNA connects a verified user and device to one named application, with no network route and nothing else visible. Applications sit behind outbound-only connectors, so they are not internet-facing and cannot be scanned by unauthenticated attackers.
For web and most TCP applications, yes. Legacy thick clients, applications that initiate connections towards the user, VoIP, and protocols requiring multicast or broadcast often need testing and sometimes a residual VPN path. Plan to run both during migration, and set a decommissioning date for the VPN so it does not persist indefinitely as an unmanaged parallel route.
No, but assurance varies sharply. Managed devices with an agent allow real posture checks on patch level, encryption, and EDR health. Unmanaged devices typically get browser-based access with limited or no posture information, since any agent reporting state runs on a machine you do not control. Set policy accordingly, granting less to unmanaged endpoints.
No. Zero trust is an architectural approach covering identity, devices, networks, applications, and data, described in NIST SP 800-207. ZTNA is one component addressing how users reach applications. Deploying ZTNA is meaningful progress, and it leaves internal east-west traffic, workload identity, and data controls to be addressed separately.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.