Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Kernel rootkits run at the same privilege level as the operating system itself, hooking system calls or editing kernel objects directly so that files, processes, and connections disappear.

What Are Kernel Rootkit Techniques?
A kernel rootkit is malicious code that executes in the same privileged context as the operating system kernel. Because every answer a security tool receives—the list of running processes, the contents of a directory, the set of open sockets—comes from the kernel, code that lives there can simply lie. Unlike a user-mode rootkit, which can be caught by comparing its output against the kernel’s own view, a kernel rootkit controls the reference point itself.
The hard part is not the hiding; it is arriving. On Linux, the usual route is a loadable kernel module installed by an attacker who already has root, or a local privilege escalation such as Dirty Pipe used as a stepping stone. On 64-bit Windows, Kernel Patch Protection and Driver Signature Enforcement mean an unsigned driver will not load at all, so attackers lean on bring-your-own-vulnerable-driver: install a legitimately signed but flawed driver—Microsoft’s vulnerable driver blocklist now names hundreds—and use its arbitrary read and write primitives to reach kernel memory without writing a driver of their own.
The classic approach replaces entries in the system call table so that getdents64 omits certain filenames, read filters lines out of /proc/net/tcp, and kill becomes a covert command channel. On modern Linux the syscall table is read-only, so rootkits either disable write protection through the CR0 register or use the ftrace and kprobes infrastructure, which is designed for exactly this kind of interception and leaves fewer obvious artefacts.
DKOM skips the function tables and edits the data structures. On Windows, unlinking an EPROCESS entry from the ActiveProcessLinks doubly linked list removes a process from every enumeration API while the scheduler, which uses a different structure, keeps running its threads. The Linux equivalent unlinks a task_struct from init_task. Nothing is hooked, so integrity checks on code pages find nothing wrong.
Rather than changing a pointer, the rootkit overwrites the first instructions of a kernel function with a jump to its own code. Variants target the interrupt descriptor table, virtual file system operation structures, or the netfilter hooks that see every packet.
Bootkits load before the operating system and therefore before its defences. LoJax, documented by ESET in 2018 and attributed to Sednit, was the first UEFI rootkit confirmed in the wild; BlackLotus, analysed in 2023, could bypass Secure Boot on fully patched Windows 11 by exploiting CVE-2022-21894. Firmware implants survive disk replacement entirely.
Turla’s Uroburos operated for years as a Windows driver with its own encrypted virtual file system. Drovorub, described in a joint NSA and FBI advisory in August 2020 and attributed to the GRU’s 85th GTsSS, combined a Linux kernel module with a client, server, and port-forwarding agent. Syslogk, found in 2022, was a Linux module that kept its payload dormant until it received a magic packet. On the criminal side, ZeroAccess and Necurs used kernel drivers to protect their own processes from removal.
Live inspection of a rootkitted host is unreliable by definition, so the useful techniques come from outside or underneath:
CAP_SYS_MODULE.Recovery is rebuild, not clean. Once arbitrary code has executed in ring 0 there is no reliable way to prove the system is trustworthy again. Our Investigation and Incident Response Services use memory acquisition and offline analysis to establish what a kernel implant did before that rebuild happens.
Sometimes, and mostly by catching the installation rather than the implant. Once the rootkit holds kernel privileges it runs at the same level as the security product and can hide files, intercept the scanner’s queries, or unload its driver. Detection usually comes from secondary evidence: a blocked vulnerable driver load, unexpected network flows seen on the switch, or memory analysis performed offline.
Bring-your-own-vulnerable-driver. Rather than trying to defeat driver signature enforcement, the attacker installs a genuine, correctly signed driver that contains a known flaw, then abuses it to read and write kernel memory. Ransomware groups have used this pattern repeatedly to disable endpoint protection. Microsoft maintains a blocklist of such drivers, which is enabled by default on recent Windows versions.
No. Code that ran with kernel privileges could have altered any part of the system, including the tools used to verify it, and firmware-level implants survive a disk wipe. The correct response is to capture memory and disk images for investigation, then rebuild from known-good media, reflash firmware where a bootkit is suspected, and rotate every credential the machine stored.
Yes. Linux is the standard platform for internet-facing services, and loadable kernel module rootkits such as Drovorub, Syslogk, and Reptile have all been found in production environments. Practical controls include Secure Boot with signed module enforcement, restricting the module-loading capability, enabling kernel lockdown mode, and shipping audit logs off the host so a local implant cannot edit them.
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.