Linux rootkits are old, but they never really disappeared. They just stopped attracting the same attention.
Most malware wants to execute and spread. A rootkit wants to stay invisible long enough that nobody notices the compromise until the damage is already done, and on Linux systems, that still works more often than many teams are comfortable admitting.
That difference matters because modern infrastructure runs heavily on Linux underneath. Cloud workloads, ESXi hosts, telecom appliances, Kubernetes nodes, internal web services. A quiet implant inside one of those systems can sit there for weeks while logs look normal and monitoring dashboards stay green.
Researchers continue seeing Linux rootkits in campaigns tied to VMware exploitation, cloud intrusions, credential theft operations, and long-term persistence activity inside enterprise environments. The tooling changes. The goal usually does not.
One of the more visible examples came from the UNC3886 activity targeting VMware environments. Researchers observed the group deploying Linux rootkits, including REPTILE and MEDUSA, after exploiting vCenter and ESXi vulnerabilities. The implants helped hide attacker activity, maintain persistence, and support credential theft across compromised systems.
The visibility problem sits underneath all of this.
Traditional security tooling still leans heavily on signatures, known malware hashes, or suspicious binaries sitting on disk. Rootkits work around that model by interfering with the operating system itself. Some operate in user space through shared libraries and linker abuse. Others load directly into the Linux kernel, where they can intercept activity before monitoring tools ever see the real picture.
A Linux rootkit exists for one reason: to make attacker activity disappear.
Sometimes that means hiding files or processes. Other times it means suppressing logs, concealing outbound connections, or masking remote access entirely. The goal is always the same — make the compromised system look normal long enough for nobody to investigate it closely.
Some rootkits stay in the user space and tamper with what applications report back to administrators. Others run directly inside the Linux kernel itself. That is where things get dangerous fast.
Once attackers reach kernel space, they can interfere with the operating system before tools like ps, top, or netstat ever see the real activity. At that point, defenders are no longer investigating a trustworthy system. They are investigating a system that may already be lying to them.
Linux powers the backbone of modern enterprise infrastructure. Many of these systems—like storage appliances or CI/CD servers—stay online for months with limited downtime, making them perfect candidates for long-term persistence.
Attackers have shifted away from older, noisy binary-replacement methods to abusing legitimate system functionality:

The line between "necessary admin tool" and "malicious rootkit" is now paper-thin.
If you suspect a system is compromised, stop relying on standard binary output. Use these commands to look for the "truth" underneath:
Target | Command / Check | Why it matters |
Modules | lsmod | Look for unsigned or unexplained modules. |
Libraries | cat /etc/ld.so.preload | This file should almost always be empty. |
eBPF | bpftool prog show | Identify hidden/unknown eBPF programs running in the kernel. |
Network | ss -antp vs netstat | Compare different tools; if they disagree, something is intercepting calls. |
Integrity | ausearch -k module-load | Check logs for unexpected module loading activity. |
Many security products still rely on static indicators like hashes or file signatures. Elastic Security Labs found that adding a single "null byte" to a rootkit’s code can be enough to bypass antivirus software entirely without breaking the malware's function.![]()
Focus on behavior, not files:
Preventing kernel-level compromise is easier than recovering from it. Once a rootkit is in the kernel, you should treat the entire system as untrusted.
Linux rootkits are not relics—they are high-value tools for attackers who understand that modern infrastructure is often a "black box" to defenders.
A hidden implant inside a cloud host or ESXi server can provide years of access while your dashboards stay green. The moment you stop trusting the operating system’s report is the moment your investigation actually begins. Traditional antivirus is only the first layer; true defense requires behavioral monitoring, deep visibility, and the readiness to rebuild rather than "clean" a compromised system.
The hardest attacks to stop are the ones that are already running, quiet and invisible, beneath your feet.