Explore our new Cybersecurity Terms Glossary, explained by our experts.

View the Glossary View the Glossary
PRIVIA

Cybersecurity Terms Glossary

What Is a Backdoor?

A backdoor is a hidden way into a system that bypasses normal authentication—planted by malware, an intruder, or a compromised supplier—giving an attacker repeat access at will.

Illustration for the cybersecurity term Backdoor
Home / Blog / Cybersecurity Terms / What Is a Backdoor?
17 November 2026

What Is a Backdoor?

What Is a Backdoor?

A backdoor is any hidden method of getting into a system that bypasses the normal authentication and security controls. It might be a malicious program listening for commands, a web shell dropped on a server, a secret administrator account, or code deliberately slipped into legitimate software. Whatever the form, its purpose is the same: to let someone return whenever they like, without being challenged and ideally without being noticed.

How Backdoors Get There

Planted After a Breach

This is the most common case. Once attackers gain initial access—through phishing, stolen credentials, or an unpatched vulnerability—they install a backdoor so they keep access even if the original hole is closed. Typical methods include creating new accounts (MITRE ATT&CK T1136), installing services or scheduled tasks (T1543, T1053), and dropping web shells on internet-facing servers (T1505.003).

Delivered as Malware

Remote access trojans such as njRAT and Remcos give an operator full control of an infected machine: file access, screenshots, keystroke capture, and command execution. Many arrive disguised as legitimate software, which is why backdoors and the trojan horse are closely linked.

Inserted Into the Supply Chain

The most dangerous backdoors arrive through software you trust. A compromised build system or a malicious contributor can place one inside a signed, official update, which then reaches every customer automatically.

Built In by the Vendor

Hard-coded support passwords, undocumented debug interfaces, and forgotten maintenance accounts are sometimes left in products by developers. They may not be malicious in intent, but once discovered they behave exactly like a planted backdoor.

Notable Cases

YearCaseWhat happened
2015Juniper ScreenOSJuniper disclosed unauthorised code in its firewall operating system, including a hard-coded administrative password (CVE-2015-7755).
2020SolarWinds OrionThe SUNBURST backdoor was inserted into signed Orion updates, downloaded by up to 18,000 customers.
2021Microsoft ExchangeAfter the ProxyLogon flaws were exploited, web shells were found on tens of thousands of on-premises Exchange servers.
2024XZ UtilsA backdoor targeting SSH authentication (CVE-2024-3094) was found in versions 5.6.0 and 5.6.1 of a widely used compression library, planted by a contributor who had spent around two years gaining maintainers’ trust.

The XZ Utils case was caught almost by accident in March 2024, when Microsoft engineer Andres Freund investigated why SSH logins on a test system were using slightly more CPU than expected. It reached only a few development and testing releases of Linux distributions before being stopped.

What a Backdoor Gives an Attacker

A backdoor turns a one-off intrusion into lasting access. Through it, attackers can return to steal more data, deploy ransomware at a time of their choosing, sell access to other criminals, enrol the machine in a botnet, or pivot deeper into the network. Most backdoors also connect outward to command and control infrastructure, because outbound traffic is less likely to be blocked than inbound connections.

Detecting a Backdoor

Removing One Safely

Deleting the file you found is rarely enough. Attackers usually leave several footholds, and if you remove one while they are watching, they may switch to another or accelerate their plans. A sound response scopes the intrusion first, identifies every persistence mechanism, rotates all credentials the attacker could have seen, and rebuilds critical systems from trusted media where integrity cannot be guaranteed. Our incident response services handle this end to end, from forensic triage to verified eradication.

The Encryption Backdoor Debate

The word also appears in policy discussions about giving governments lawful access to encrypted data. In the UK, the Investigatory Powers Act 2016 allows the Home Secretary to issue technical capability notices. In February 2025, Apple withdrew its Advanced Data Protection feature for UK users following a reported notice of this kind. Security researchers widely argue that any deliberate access mechanism weakens security for everyone, because it can be found and abused by others.

FAQ – Frequently Asked Questions

What Is the Difference Between a Backdoor and a Trojan?

A trojan describes how malware gets in: it disguises itself as something legitimate so the user installs it. A backdoor describes what it does once present: it provides hidden, repeatable access. Many trojans install backdoors, but backdoors can also be web shells, rogue accounts, or code inserted into software without any trojan involved.

Can Antivirus Software Detect Backdoors?

It can detect many known backdoor families, but often misses custom malware, web shells written for a single target, and backdoors built from legitimate tools such as rogue accounts, SSH keys, or remote management software. Endpoint detection and response, file integrity monitoring, network monitoring, and periodic threat hunting give much better coverage than signature-based antivirus alone.

What Should I Do if I Find a Backdoor on a Server?

Do not delete it immediately. Isolate the system if you can do so safely, preserve evidence such as memory and logs, and treat it as a confirmed security incident. The backdoor is evidence of a wider intrusion, so you need to understand how the attacker got in and what else they touched before eradicating it and restoring service.

How Do Supply Chain Backdoors Get Past Code Signing?

Code signing proves that software came from the vendor’s build process and has not been altered afterwards. If attackers compromise that build process, or a trusted contributor inserts malicious code into the source, the backdoor is signed along with everything else. Reproducible builds, build system hardening, and reviewing behaviour after updates help reduce this risk.

Related Terms

You May Be Interested In These

What Is Substation Automation Security?

Cybersecurity Terms

17 November 2026

What Is Substation Automation Security?

Read More
What Is Container Registry Security?

Cybersecurity Terms

17 November 2026

What Is Container Registry Security?

Read More