Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Exposed Remote Desktop has been a leading ransomware entry point for years. Securing it means removing it from the internet, then hardening what remains.

What Is RDP Security?
Remote Desktop Protocol gives an administrator a full interactive session on a Windows machine. That is exactly what an intruder wants, which is why RDP has spent years at or near the top of initial access vectors in incident response statistics from Coveware, Sophos, and the NCSC alike. The protocol itself is not unusually weak. The problem is how it is deployed: published directly to the internet, authenticated by a password alone, on a machine that nobody patches on an emergency schedule.
Credential attacks. The common route, and it needs no vulnerability. Scanning the internet for TCP 3389 is a matter of minutes, and the results are published continuously by services such as Shodan. Attackers then try credentials harvested from infostealer logs, bought from access brokers, or guessed through password spraying against predictable accounts. Once a session opens, the attacker has an interactive desktop with whatever privileges that account holds.
Exploits. BlueKeep, CVE-2019-0708, was pre-authentication remote code execution in Remote Desktop Services on Windows 7 and Server 2008, serious enough that Microsoft issued patches for out-of-support systems in May 2019 and the NCSC published a specific alert. A related set, the DejaBlue vulnerabilities of August 2019, affected newer versions. CVE-2019-0887 allowed a malicious server to attack a connecting client through the clipboard path, which is a reminder that the risk runs in both directions.
Protocol-level interception. Without Network Level Authentication and proper certificate validation, an on-path attacker can present their own certificate and relay the session. Users click through the warning, because they always have. CredSSP had its own flaw in CVE-2018-0886, allowing credential relay to a target server.
Do not publish RDP to the internet. Every other recommendation is secondary to this one.
Remote access should terminate on something designed to face the internet — a VPN concentrator with multi-factor authentication, a Remote Desktop Gateway with MFA, or a ZTNA broker that publishes the specific desktop rather than a route into the network. Changing the listening port away from 3389 is not a control; mass scanning covers all ports and the service announces itself in the banner.
Verify this rather than assume it. Organisations that are certain RDP is not exposed frequently find a forgotten rule on a branch firewall, a cloud instance with a permissive security group, or a port forward created for a supplier in 2020.
UK cyber insurers now ask about exposed RDP explicitly in proposal forms, and several have declined or limited cover where it is present without MFA. Cyber Essentials requires that administrative access is not exposed to the internet without appropriate controls, and the NCSC’s guidance on remote access is clear on the point. If you need one argument to get the firewall rule removed, the renewal questionnaire is usually more persuasive than the threat model.
Our Infrastructure Security Testing Services enumerate what your perimeter publishes and test the authentication path behind it.
Barely. Mass scanning covers the full port range, and the service identifies itself regardless of which port it listens on. Port changes may reduce noise from the least sophisticated automated attacks, but they provide no meaningful protection against anyone actually looking. Removing internet exposure entirely is the control that matters.
No, though it is important. NLA requires authentication before a session is established, which blocks pre-authentication exploits such as BlueKeep and reduces resource exhaustion attacks. It does nothing against an attacker who has valid credentials, which is how most RDP intrusions begin. Pair it with multi-factor authentication and remove the service from the internet.
Through the same controlled path as employees: a gateway or ZTNA broker with multi-factor authentication, time-limited access to named systems only, individual accounts rather than shared ones, and session recording where the access is privileged. A permanent firewall rule allowing a supplier’s address range to reach RDP is the arrangement that keeps appearing in breach reports.
Event ID 4625 for failed logons, which reveals password spraying, and event ID 4624 with logon type 10 for successful remote interactive sessions. Alert on logons from unfamiliar source addresses, outside working hours, or to systems a user does not normally access. Internally, any workstation-to-workstation RDP is worth investigating as probable lateral movement.
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.