Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
VDI security covers the controls around hosted desktops: the internet-facing gateway, the golden image, session redirection settings, and the fact that hundreds of desktops now share one set of servers.

What Is Virtual Desktop Infrastructure Security?
Virtual desktop infrastructure moves the desktop into the data centre or the cloud and leaves the user with a thin client and a session. The security case is genuine: data never lands on the endpoint, a lost laptop carries nothing, and contractors and offshore teams can be given a controlled environment rather than a build you cannot manage. The security consequences are equally real, because the controls move with the desktop and the exposure concentrates in a small number of components.
Remote access is the point of VDI, so there is an internet-facing component, and it has repeatedly been the way in. Citrix Bleed, CVE-2023-4966, leaked session tokens from NetScaler appliances and allowed attackers to resume authenticated sessions while bypassing multi-factor authentication entirely; it was used in ransomware campaigns through late 2023 and into 2024.
VMware Horizon servers running vulnerable Log4j were exploited at scale from January 2022, and the NCSC issued UK-specific guidance after the technique was used against health and public sector targets. Earlier Citrix ADC vulnerabilities, including CVE-2019-19781, produced the same pattern.
Three conclusions follow. Gateway appliances belong in the fastest patch tier you have, ahead of ordinary servers. Multi-factor authentication must be enforced at the gateway rather than inside the session, though token theft shows that even this is not absolute. And when a gateway vulnerability involving session tokens is disclosed, terminate all active sessions after patching, because patching alone leaves stolen tokens valid.
In a non-persistent deployment, every user receives a copy of one golden image. That is an advantage, because a compromise within a session disappears at logoff, and a liability, because a weakness in the image is replicated several hundred times and a compromise of the image pipeline reaches every desktop at once.
Treat image building as a controlled release process: hardened baseline, current patches at the point of publication, no embedded credentials or cached tokens, approved software only, and restricted access to the management console that publishes it. Audit for secrets left in the image, which in our reviews is where service account passwords and API keys most often turn up.
Most of the data loss risk in VDI sits in redirection, and the defaults are set for convenience.
Published applications must not become published desktops. Any dialogue that can reach a file browser, a help window with a link, or a print-to-file prompt can often be escalated into a shell, which is a standard finding in VDI testing. Restrict the shell, apply application allowlisting inside the image, and test the published application set directly rather than assuming the publishing mechanism contains it.
Remember also that desktops share infrastructure. Users sitting on the same host and often the same network segment make lateral movement between sessions straightforward unless host firewalls block peer connections and sessions are segmented. And in a non-persistent environment, forensic evidence evaporates at logoff, so EDR telemetry and event logs must be forwarded off the session in real time or an investigation will have nothing to examine.
Patch the gateway first, enforce multi-factor authentication and revoke sessions after gateway incidents, lock down redirection, harden and audit the golden image, deploy EDR inside the image with central log forwarding, and segment session hosts from each other. For the hypervisor layer underneath, see hypervisor security.
At Privia Security, we test VDI as an environment rather than as a product: gateway, session, image, and the path between a published application and a command prompt. Our Penetration Test Services include breakout testing from published desktops and applications.
It is different rather than uniformly better. Data stays in the data centre, lost devices carry nothing, and the build is centrally controlled, which suits contractors and third parties. In exchange, risk concentrates in the gateway and the golden image, and a single vulnerability in either affects everyone at once. The security benefit depends on patch discipline and redirection settings rather than on the architecture itself.
It leaked session tokens from memory on NetScaler appliances, letting attackers resume an authenticated session without credentials and without satisfying multi-factor authentication. Patching did not invalidate tokens already stolen, so organisations that applied the update without terminating all active sessions remained exposed. Several ransomware incidents in late 2023 began exactly that way.
Frequently, if the application set has not been tested. Any route to a file open dialogue, an embedded browser, a help window, or a print-to-file prompt can often be turned into a file manager and then a command shell. Restricting the shell, applying application allowlisting inside the image, and testing each published application explicitly are the controls that close it.
Only with telemetry collected while the session was live. Non-persistent desktops are destroyed at logoff, so local artefacts are gone before anyone looks. Deploy EDR inside the golden image, forward Windows event logs and process telemetry to a central platform in real time, and define a procedure to snapshot a suspect session before it is torn down.
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.