Kernel hardening is the set of defensive changes that make an operating system kernel harder to abuse when a bug exists. It does not remove vulnerabilities; instead, it reduces the chance that a memory corruption, logic flaw, or race condition can be turned into code execution, privilege escalation, or a sandbox escape.
In practice, hardening includes attack-surface reduction, syscall filtering, memory protections, control-flow protections, and stricter configuration of kernel features and modules. These measures matter because the kernel sits close to the trust boundary: if an attacker can influence it, they may gain control over the host, weaken isolation, or pivot from a guest environment toward a broader compromise. Defenders use hardening to raise exploit cost and limit blast radius while they patch the underlying flaw.